Rheology of Fresh Cement-Based Materials: Fundamentals, Measurements, and Applications 1032208015, 9781032208015

This book introduces fundamentals, measurements, and applications of rheology of fresh cement-based materials. The rheol

176 3 41MB

English Pages 338 [339] Year 2023

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Cover
Half Title
Title Page
Copyright Page
Table of Contents
Preface
Authors
1 Introduction to rheology
1.1 The Subject and Object of Rheology
1.2 Basic Principles of Rheology
1.2.1 Definition of viscosity
1.2.2 Newtonian flow
1.2.3 Non-Newtonian flow
1.2.4 Thixotropy
1.2.5 Anti-thixotropy (rheopexy)
1.3 Cement-Based Materials
1.3.1 History of cement and concrete
1.3.2 Fresh properties of cement-based materials
1.4 The Scope of This Book
References
2 Rheology for cement paste
2.1 Interaction between Particles in the Paste
2.1.1 Colloidal interaction
2.1.1.1 Van der Waals force
2.1.1.2 Electrostatic repulsion
2.1.1.3 Steric hinder force
2.1.2 Brownian forces
2.1.3 Hydrodynamic force
2.2 Effect of Compositions on Rheology
2.2.1 Volume fraction
2.2.2 Interstitial solution
2.2.3 Cement
2.2.4 Mineral admixture
2.2.4.1 Fly ash
2.2.4.2 Ground blast furnace slag
2.2.4.3 Silica fume
2.2.4.4 Limestone powder
2.2.4.5 Ternary binder system
2.2.5 Chemical admixtures
2.2.5.1 Superplasticizer
2.2.5.2 Viscosity-modifying agent
2.2.5.3 Air-entraining agent
2.3 Effect of Temperature on Rheology
2.4 Effect of Shearing on Rheology
2.5 Effect of Pressure on Rheology
2.6 Summary
References
3 Rheological properties of fresh concrete materials
3.1 General Considerations for Granular Materials
3.2 Flow Regimes of Concrete
3.2.1 Relationships between aggregate volume fraction and concrete rheology
3.2.1.1 Viscosity vs aggregate volume fraction
3.2.1.2 Yield stress vs aggregate volume fraction
3.2.2 Excess paste theory
3.3 Influence of Aggregate Characteristics
3.3.1 Aggregate volume fraction
3.3.2 Gradation and particle size
3.3.3 Particle morphology
3.4 Effect of External Factors
3.4.1 Mixing process
3.4.2 Shear history
3.4.3 Measuring geometry
3.5 Summary
References
4 Empirical techniques evaluating concrete rheology
4.1 Introduction
4.2 Slump: ASTM Abrams Cone
4.2.1 Geometry
4.2.2 Testing procedure and parameters
4.2.3 Data interpretation
4.3 Slump Flow and T50
4.3.1 Geometry and testing procedure
4.3.2 Data interpretation
4.4 V-Funnel Test Flow Time
4.4.1 Geometry
4.4.2 Testing procedure
4.4.3 Data interpretation
4.5 Other Methods
4.5.1 L-box
4.5.2 LCPC box
4.5.3 V-funnel coupled with a horizontal channel
4.5.4 J-ring
4.6 Summary
References
5 Paste rheometers
5.1 Introduction to the Rheology of Cement Paste
5.2 Rheometers for Cement Paste
5.2.1 Narrow gap coaxial cylinder rheometer
5.2.1.1 Geometry
5.2.1.2 Measurement principle
5.2.1.3 Measuring errors and artifacts
5.2.2 Plate – plate rheometer
5.2.2.1 Geometry
5.2.2.2 Measurement principle
5.2.2.3 Measuring errors and artifacts
5.2.3 Other rheometers
5.2.3.1 Capillary viscometer
5.2.3.2 Falling sphere viscometer
5.3 Measuring Procedures
5.3.1 Flow curves test
5.3.2 Static yield stress test
5.3.3 Oscillatory shear test
5.3.3.1 Description of SAOS and LAOS
5.3.3.2 Measurement principle
5.3.3.3 Application to cement paste
5.4 Summary
References
6 Concrete rheometers
6.1 Introduction
6.2 Tests Methods and Principles
6.2.1 Coaxial cylinder rheometer
6.2.1.1 Searle rheometer
6.2.1.2 Couette rheometer
6.2.1.3 Principle
6.2.1.4 Measuring errors and artifacts
6.2.2 Parallel-plate rheometer
6.2.2.1 Geometry
6.2.2.2 Principle
6.2.2.3 Measuring errors and artifacts
6.2.3 Other rheometers
6.2.3.1 CEMAGREF-IMG rheometer
6.2.3.2 Viskomat XL
6.2.3.3 The IBB rheometer
6.2.3.4 Rheometer developed in China
6.2.3.5 The modifications of the BTRHEOM rheometer
6.2.3.6 Other instruments
6.3 Measuring Procedures
6.3.1 Preparation of specimen
6.3.2 The testing procedures of ICAR
6.3.3 The testing procedures of ConTec Viscometer 5
6.3.4 The testing procedures of the BTRHEOM rheometer
6.4 Data Collection and Processing
6.4.1 Static yield stress test
6.4.2 The flow curve test
6.4.3 Thixotropy test
6.5 Relation of Rheological Parameters Measured by Different Rheometers
6.6 Summary
References
7 Mixture design of concrete based on rheology
7.1 Introduction
7.2 Principles of Mixture Design Methods Based on Rheology
7.2.1 Vectorized-rheograph approach
7.2.2 Paste rheology criteria
7.2.3 Concrete rheology method
7.2.4 Excess paste theory
7.2.5 Simplex centroid design method
7.3 Typical Examples of Mixture Design
7.3.1 Paste rheology criteria proposed by Wu and An
7.3.2 Paste rheology model proposed by Ferrara et al.
7.3.3 Concrete rheology method of Abo Dhaheer et al.
7.3.4 Simplex centroid design method proposed by Jiao et al.
7.4 Summary
References
8 Rheology and self-compacting concrete
8.1 Introduction to SCC
8.1.1 Brief history of SCC
8.1.2 Raw materials of SCC
8.1.2.1 Powder
8.1.2.2 Chemical admixtures
8.1.2.3 Aggregates
8.1.2.4 Water
8.1.3 Mix proportion of SCC
8.1.3.1 Laboratory experiments and empirical parameters
8.1.3.2 Statistical method
8.1.3.3 Maximum packing density
8.1.3.4 Other methods
8.1.4 Application of SCC
8.2 Rheology of SCC
8.2.1 Factors affecting rheology of SCC
8.2.1.1 Fly ash
8.2.1.2 Rice husk ash
8.2.1.3 Silica fume
8.2.1.4 Metakaolin
8.2.1.5 Blast furnace slag
8.2.1.6 Fibers
8.2.1.7 Air-entraining agent
8.2.1.8 Superplasticizer
8.2.1.9 Recycled concrete aggregates
8.2.1.10 Binary and ternary binder system
8.2.1.11 Other constituents
8.2.2 Special rheological behaviors
8.2.2.1 Thixotropy
8.2.2.2 Shear-thinning or shear-thickening behavior
8.3 Formwork Pressure of SCC
8.3.1 Factors affecting formwork pressure
8.3.2 Formwork pressure prediction
8.3.2.1 Method proposed by Gardner (Gardner et al., 2012)
8.3.2.2 Method proposed by Khayat (Khayat and Omran, 2010)
8.4 Stability of SCC
8.4.1 Static stability
8.4.2 Dynamic stability
8.5 Summary
References
9 Rheology of other cement-based materials
9.1 Rheology of Alkali-Activated Materials (AAMs)
9.1.1 Introduction
9.1.2 Effect of alkaline activators on rheology of AAMs
9.1.2.1 Na/KOH
9.1.2.2 Na/K-silicates
9.1.3 Effect of precursors on the rheology of AAMs
9.1.3.1 Chemical and physical properties of precursors
9.1.4 Effects of chemical admixtures on the rheology of AAMs
9.1.4.1 Water-reducing admixtures
9.1.4.2 Other chemical admixtures
9.1.5 Effects of mineral additions on the rheology of AAMs
9.1.5.1 Reactive mineral additions
9.1.5.2 Inert mineral additions
9.1.6 Effect of aggregates on the rheology of AAMs
9.2 Rheology of Cement Paste Backfilling (CPB)
9.2.1 Introduction
9.2.2 Factors affecting the rheological properties of CPB
9.2.2.1 Cement
9.2.2.2 Solid concentration
9.2.2.3 Mixing intensity
9.2.2.4 Particle size
9.2.2.5 High-range water reducer (HRWR)
9.2.2.6 Temperature
9.2.2.7 Other constituents
9.3 Rheology of Fiber-Reinforced, Cement-Based Materials
9.3.1 Introduction
9.3.2 Influence of fiber on the rheology of FRCs
9.3.2.1 Fiber orientation
9.3.2.2 Fiber length
9.3.2.3 Types of fiber
9.3.3 Effect of fibers on the rheology of AAMs
9.3.4 Prediction of the yield stress of FRC
9.3.5 Prediction of plastic viscosity
9.4 Summary
9.4.1 AAMs
9.4.2 Cement paste backfilling
9.4.3 Fiber-reinforced, cement-based materials
References
10 Rheology and Pumping
10.1 Introduction
10.2 Characterization of Pumpability
10.2.1 Definition of pumpability
10.2.2 Determination of pumpability
10.3 Lubrication Layer
10.3.1 The formation of the lubrication layer
10.3.2 The determination of the lubrication layer
10.3.2.1 Tribometer
10.3.2.2 Sliding pipe
10.3.2.3 Other methods
10.4 Prediction of Pumping
10.4.1 Empirical model for pumping prediction
10.4.2 Numerical model for pumping prediction
10.4.3 Computer simulations for pressure loss predictions
10.5 Effect of Pumping on the Fresh Properties of Concrete
10.5.1 Air content
10.5.2 Rheology
10.5.2.1 Yield stress and plastic viscosity
10.5.2.2 Thixotropy
10.5.2.3 Supplementary cementitious materials (SCMs)
10.5.2.4 Water absorption by aggregates
10.5.2.5 Water-reducing agent
10.6 Summary
References
11 Rheology and 3D printing
11.1 Introduction to 3D-Printing Concrete
11.1.1 Development of 3D printing technology
11.1.2 Requirement for 3D printing concrete
11.2 Printability of 3D Printing Concrete
11.2.1 The definition of printability
11.2.2 The test for printability
11.2.3 Criteria to evaluate the loading bearing capacity
11.2.4 Printable cement-based materials
11.3 Interlayer Bonding and Rheology
11.3.1 The characterization of interlayer bonding
11.3.2 The effect of rheological properties on interlayer bonding
11.3.3 The effect of rheological properties on interface durability
11.3.4 The effect of rheological properties on interface microstructure
11.4 Summary
References
Index
Recommend Papers

Rheology of Fresh Cement-Based Materials: Fundamentals, Measurements, and Applications
 1032208015, 9781032208015

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Rheology of Fresh Cement-Based Materials This book introduces fundamentals, measurements, and applications of rheology of fresh cement-based materials. The rheology of a fresh cement-based material is one of its most important aspects, characterizing its flow and deformation, and governing the mixing, placement, and casting quality of a concrete. This is the first book that brings the field together on an increasingly important topic, as new types of cement-based materials and new concrete technologies are developed. It describes measurement equipment, procedures, and data interpretation of the rheology of cement paste and concrete, as well as applications such as self-compacting concrete, pumping, and 3D printing. A range of other cement-based materials such as fiber-reinforced concrete, cemented paste backfills, and alkali-activated cement are also examined. Rheology of Fresh Cement-Based Materials serves as a reference book for researchers and engineers, and a textbook for advanced undergraduate and graduate students.

Rheology of Fresh Cement-Based Materials Fundamentals, Measurements, and Applications

Qiang Yuan, Caijun Shi, and Dengwu Jiao

MATLAB ® is a trademark of The MathWorks, Inc. and is used with permission. The MathWorks does not warrant the accuracy of the text or exercises in this book. This book’s use or discussion of MATLAB ® software or related products does not constitute endorsement or sponsorship by The MathWorks of a particular pedagogical approach or particular use of the MATLAB ® software.

First edition published 2023 by CRC Press 4 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN and by CRC Press 6000 Broken Sound Parkway NW, Suite 300, Boca Raton, FL 33487-2742 © 2023 Qiang Yuan, Caijun Shi and Dengwu Jiao CRC Press is an imprint of Informa UK Limited The right of Qiang Yuan, Caijun Shi and Dengwu Jiao to be identified as authors of this work has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. For permission to photocopy or use material electronically from this work, access www.copyright.com or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. For works that are not available on CCC please contact [email protected] Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Names: Yuan, Qiang, author. | Shi, Caijun, author. | Jiao, Dengwu, author. Title: Rheology of fresh cement-based materials : fundamentals, measurements, and applications / Qiang Yuan, Caijun Shi, and Dengwu Jiao. Description: First edition. | Boca Raton : CRC Press, 2023. | Includes bibliographical references and index. Identifiers: LCCN 2022033778 | ISBN 9781032208015 (hbk) | ISBN 9781032208022 (pbk) | ISBN 9781003265313 (ebk) Subjects: LCSH: Concrete—Viscosity. | Concrete—Mixing. | Rheology. Classification: LCC TA440 .Y833 2023 | DDC 620.1/36—dc23/eng/20221011 LC record available at https://lccn.loc.gov/2022033778 ISBN: 978-1-032-20801-5 (hbk) ISBN: 978-1-032-20802-2 (pbk) ISBN: 978-1-003-26531-3 (ebk) DOI: 10.1201/9781003265313 Typeset in Sabon by codeMantra

Contents

Preface Authors 1 Introduction to rheology

xiii xv 1

1.1 1.2

The Subject and Object of Rheology 1 Basic Principles of Rheology 7 1.2.1 Definition of viscosity 7 1.2.2 Newtonian flow 8 1.2.3 Non-Newtonian flow 9 1.2.4 Thixotropy 11 1.2.5 Anti-thixotropy (rheopexy) 13 1.3 Cement-Based Materials 13 1.3.1 History of cement and concrete 13 1.3.2 Fresh properties of cement-based materials 16 1.4 The Scope of This Book 20 References 21

2 Rheology for cement paste 2.1

2.2

25

Interaction between Particles in the Paste 25 2.1.1 Colloidal interaction 25 2.1.1.1 Van der Waals force 26 2.1.1.2 Electrostatic repulsion 26 2.1.1.3 Steric hinder force 26 2.1.2 Brownian forces 26 2.1.3 Hydrodynamic force 27 Effect of Compositions on Rheology 27 2.2.1 Volume fraction 27 2.2.2 Interstitial solution 29 2.2.3 Cement 30 2.2.4 Mineral admixture 31 2.2.4.1 Fly ash 31 2.2.4.2 Ground blast furnace slag 33 2.2.4.3 Silica fume 34

v

vi

Contents

2.2.4.4 Limestone powder 36 2.2.4.5 Ternary binder system 37 2.2.5 Chemical admixtures 38 2.2.5.1 Superplasticizer 38 2.2.5.2 Viscosity-modifying agent 39 2.2.5.3 Air-entraining agent 40 2.3 Effect of Temperature on Rheology 41 2.4 Effect of Shearing on Rheology 41 2.5 Effect of Pressure on Rheology 42 2.6 Summary 43 References 43

3 Rheological properties of fresh concrete materials

51

3.1 3.2

General Considerations for Granular Materials 51 Flow Regimes of Concrete 51 3.2.1 Relationships between aggregate volume fraction and concrete rheology 52 3.2.1.1 Viscosity vs aggregate volume fraction 52 3.2.1.2 Yield stress vs aggregate volume fraction 53 3.2.2 Excess paste theory 56 3.3 Influence of Aggregate Characteristics 59 3.3.1 Aggregate volume fraction 59 3.3.2 Gradation and particle size 60 3.3.3 Particle morphology 62 3.4 Effect of External Factors 67 3.4.1 Mixing process 67 3.4.2 Shear history 69 3.4.3 Measuring geometry 69 3.5 Summary 69 References 70

4 Empirical techniques evaluating concrete rheology 4.1 4.2

4.3

4.4

Introduction 75 Slump: ASTM Abrams Cone 75 4.2.1 Geometry 76 4.2.2 Testing procedure and parameters 76 4.2.3 Data interpretation 77 Slump Flow and T50 80 4.3.1 Geometry and testing procedure 81 4.3.2 Data interpretation 82 V-Funnel Test Flow Time 82 4.4.1 Geometry 83 4.4.2 Testing procedure 84 4.4.3 Data interpretation 84

75

Contents

vii

4.5

Other Methods 85 4.5.1 L-box 85 4.5.2 LCPC box 87 4.5.3 V-funnel coupled with a horizontal channel 88 4.5.4 J-ring 89 4.6 Summary 91 References 91

5 Paste rheometers

95

5.1 5.2

Introduction to the Rheology of Cement Paste 95 Rheometers for Cement Paste 96 5.2.1 Narrow gap coaxial cylinder rheometer 96 5.2.1.1 Geometry 96 5.2.1.2 Measurement principle 96 5.2.1.3 Measuring errors and artifacts 98 5.2.2 Plate–plate rheometer 99 5.2.2.1 Geometry 99 5.2.2.2 Measurement principle 101 5.2.2.3 Measuring errors and artifacts 102 5.2.3 Other rheometers 104 5.2.3.1 Capillary viscometer 104 5.2.3.2 Falling sphere viscometer 107 5.3 Measuring Procedures 109 5.3.1 Flow curves test 109 5.3.2 Static yield stress test 110 5.3.3 Oscillatory shear test 111 5.3.3.1 Description of SAOS and LAOS 111 5.3.3.2 Measurement principle 112 5.3.3.3 Application to cement paste 115 5.4 Summary 117 References 117

6 Concrete rheometers 6.1 6.2

Introduction 123 Tests Methods and Principles 124 6.2.1 Coaxial cylinder rheometer 125 6.2.1.1 Searle rheometer 126 6.2.1.2 Couette rheometer 126 6.2.1.3 Principle 128 6.2.1.4 Measuring errors and artifacts 131 6.2.2 Parallel-plate rheometer 137 6.2.2.1 Geometry 137 6.2.2.2 Principle 138 6.2.2.3 Measuring errors and artifacts 139

123

viii

Contents

6.2.3

Other rheometers 140 6.2.3.1 CEMAGREF-IMG rheometer 140 6.2.3.2 Viskomat XL 140 6.2.3.3 The IBB rheometer 140 6.2.3.4 Rheometer developed in China 142 6.2.3.5 The modifications of the BTRHEOM rheometer 142 6.2.3.6 Other instruments 143 6.3 Measuring Procedures 143 6.3.1 Preparation of specimen 143 6.3.2 The testing procedures of ICAR 143 6.3.3 The testing procedures of ConTec Viscometer 5 145 6.3.4 The testing procedures of the BTRHEOM rheometer 147 6.4 Data Collection and Processing 148 6.4.1 Static yield stress test 148 6.4.2 The flow curve test 149 6.4.3 Thixotropy test 150 6.5 Relation of Rheological Parameters Measured by Different Rheometers 150 6.6 Summary 152 References 152

7 Mixture design of concrete based on rheology

155

7.1 7.2

Introduction 155 Principles of Mixture Design Methods Based on Rheology 156 7.2.1 Vectorized-rheograph approach 156 7.2.2 Paste rheology criteria 158 7.2.3 Concrete rheology method 164 7.2.4 Excess paste theory 166 7.2.5 Simplex centroid design method 167 7.3 Typical Examples of Mixture Design 169 7.3.1 Paste rheology criteria proposed by Wu and An 169 7.3.2 Paste rheology model proposed by Ferrara et al. 169 7.3.3 Concrete rheology method of Abo Dhaheer et al. 173 7.3.4 Simplex centroid design method proposed by Jiao et al. 176 7.4 Summary 178 References 178

8 Rheology and self-compacting concrete 8.1

Introduction to SCC 183 8.1.1 Brief history of SCC 183 8.1.2 Raw materials of SCC 184 8.1.2.1 Powder 184 8.1.2.2 Chemical admixtures 185 8.1.2.3 Aggregates 185 8.1.2.4 Water 186

183

Contents

ix

8.1.3

Mix proportion of SCC 186 8.1.3.1 Laboratory experiments and empirical parameters 186 8.1.3.2 Statistical method 187 8.1.3.3 Maximum packing density 187 8.1.3.4 Other methods 188 8.1.4 Application of SCC 188 8.2 Rheology of SCC 190 8.2.1 Factors affecting rheology of SCC 190 8.2.1.1 Fly ash 190 8.2.1.2 Rice husk ash 190 8.2.1.3 Silica fume 191 8.2.1.4 Metakaolin 192 8.2.1.5 Blast furnace slag 192 8.2.1.6 Fibers 192 8.2.1.7 Air-entraining agent 193 8.2.1.8 Superplasticizer 193 8.2.1.9 Recycled concrete aggregates 195 8.2.1.10 Binary and ternary binder system 197 8.2.1.11 Other constituents 197 8.2.2 Special rheological behaviors 198 8.2.2.1 Thixotropy 198 8.2.2.2 Shear-thinning or shear-thickening behavior 199 8.3 Formwork Pressure of SCC 200 8.3.1 Factors affecting formwork pressure 201 8.3.2 Formwork pressure prediction 201 8.3.2.1 Method proposed by Gardner (Gardner et al., 2012) 202 8.3.2.2 Method proposed by Khayat (Khayat and Omran, 2010) 202 8.4 Stability of SCC 204 8.4.1 Static stability 204 8.4.2 Dynamic stability 206 8.5 Summary 208 References 209

9 Rheology of other cement-based materials 9.1

Rheology of Alkali-Activated Materials (AAMs) 215 9.1.1 Introduction 215 9.1.2 Effect of alkaline activators on rheology of AAMs 217 9.1.2.1 Na/KOH 217 9.1.2.2 Na/K-silicates 219 9.1.3 Effect of precursors on the rheology of AAMs 223 9.1.3.1 Chemical and physical properties of precursors 223 9.1.4 Effects of chemical admixtures on the rheology of AAMs 225 9.1.4.1 Water-reducing admixtures 225 9.1.4.2 Other chemical admixtures 231

215

x

Contents

9.1.5

Effects of mineral additions on the rheology of AAMs 232 9.1.5.1 Reactive mineral additions 232 9.1.5.2 Inert mineral additions 233 9.1.6 Effect of aggregates on the rheology of AAMs 233 9.2 Rheology of Cement Paste Backfilling (CPB) 234 9.2.1 Introduction 234 9.2.2 Factors affecting the rheological properties of CPB 235 9.2.2.1 Cement 235 9.2.2.2 Solid concentration 236 9.2.2.3 Mixing intensity 237 9.2.2.4 Particle size 237 9.2.2.5 High-range water reducer (HRWR) 238 9.2.2.6 Temperature 239 9.2.2.7 Other constituents 239 9.3 Rheology of Fiber-Reinforced, Cement-Based Materials 240 9.3.1 Introduction 240 9.3.2 Influence of fiber on the rheology of FRCs 242 9.3.2.1 Fiber orientation 242 9.3.2.2 Fiber length 243 9.3.2.3 Types of fiber 244 9.3.3 Effect of fibers on the rheology of AAMs 246 9.3.4 Prediction of the yield stress of FRC 248 9.3.5 Prediction of plastic viscosity 248 9.4 Summary 249 9.4.1 AAMs 249 9.4.2 Cement paste backfilling 250 9.4.3 Fiber-reinforced, cement-based materials 251 References 251

10 Rheology and Pumping 10.1 Introduction 259 10.2 Characterization of Pumpability 260 10.2.1 Definition of pumpability 260 10.2.2 Determination of pumpability 261 10.3 Lubrication Layer 261 10.3.1 The formation of the lubrication layer 261 10.3.2 The determination of the lubrication layer 262 10.3.2.1 Tribometer 262 10.3.2.2 Sliding pipe 263 10.3.2.3 Other methods 264 10.4 Prediction of Pumping 264 10.4.1 Empirical model for pumping prediction 265 10.4.2 Numerical model for pumping prediction 266 10.4.3 Computer simulations for pressure loss predictions 267

259

Contents

xi

10.5 Effect of Pumping on the Fresh Properties of Concrete 268 10.5.1 Air content 268 10.5.2 Rheology 273 10.5.2.1 Yield stress and plastic viscosity 273 10.5.2.2 Thixotropy 278 10.5.2.3 Supplementary cementitious materials (SCMs) 281 10.5.2.4 Water absorption by aggregates 283 10.5.2.5 Water-reducing agent 284 10.6 Summary 284 References 285

11 Rheology and 3D printing

291

11.1 Introduction to 3D-Printing Concrete 291 11.1.1 Development of 3D printing technology 291 11.1.2 Requirement for 3D printing concrete 295 11.2 Printability of 3D Printing Concrete 296 11.2.1 The definition of printability 296 11.2.2 The test for printability 297 11.2.3 Criteria to evaluate the loading bearing capacity 299 11.2.4 Printable cement-based materials 301 11.3 Interlayer Bonding and Rheology 303 11.3.1 The characterization of interlayer bonding 303 11.3.2 The effect of rheological properties on interlayer bonding 305 11.3.3 The effect of rheological properties on interface durability 308 11.3.4 The effect of rheological properties on interface microstructure 309 11.4 Summary 311 References 311

Index

317

Preface

Rheology, the study of the flow and deformation of matter, is an independent and prominent branch of the natural sciences. It is an extremely useful technical tool and a scientific basis for cementitious materials. Advanced technologies such as pumping and 3D printing impose additional requirements on the rheological properties of cement-based materials. Besides, rheology for cement-based materials extends beyond the study of flow and deformation. The significance of the physical and chemical reactions underlies the growing importance of rheology. This intricate and intriguing rheology of cement-based materials attracts the interest of an increasing number of researchers. A lot of research has been done in the rheology of cement-based materials, encompassing its fundamentals, measurements, and predictions. This does not imply, however, that we have a complete grasp of the rheology of cementbased materials and can accurately predict as well as control it. Still, many scientific issues are yet to be clarified. This book focuses on the flow and rheological behavior of fresh cement-based materials. It reviews the fundamentals, measurement techniques, and applications involved in these areas, and presents the state-of-the-art progress made recently. The authors of this book have long been involved in research pertaining to the rheology of fresh cement-based materials at Central South University, Hunan University, and Ghent University. The essential sources for this book were the graduate students who obtained their M.Sc. and Ph.D. degrees in this field of study in the past few years. Obviously, this book could not have been completed without their acquired knowledge, experience, and outstanding contributions. The authors would like to express their gratitude to everyone who contributed to the content of this book. They also gratefully recognize the contributions of reviewers and students who assisted in organizing the content of the entire book into the present layout, as well as those who drew the updated graphs and tables. The intended audience of this book includes students, researchers, and engineers in concrete technology. Researchers may be inspired by the comprehensive overview of the rheology of cement and concrete. Practicing engineers can also benefit from this book by imparting fundamental knowledge and practical techniques. Qiang Yuan Caijun Shi Dengwu Jiao

xiii

xiv Preface

MATLAB® is a registered trademark of The MathWorks, Inc. For product information, please contact: The MathWorks, Inc. 3 Apple Hill Drive Natick, MA 01760-2098 USA Tel: 508-647-7000 Fax: 508-647-7001 E-mail: [email protected] Web: www.mathworks.com

Authors

Qiang Yuan is a professor in the School of Civil Engineering at Central South University, China. He is the author of Transport and Interactions of Chlorides in Cement-Based Materials, also published by CRC Press. Caijun Shi is a professor in the College of Civil Engineering at Hunan University, China. He is the author of Alkali-Activated and Concretes and Transport and Interactions of Chlorides in Cement-Based Materials, also published by CRC Press. Dengwu Jiao is a postdoctoral researcher in the Department of Structural Engineering and Building Materials at Ghent University.

xv

Chapter 1

Introduction to rheology

1.1 THE SUBJECT AND OBJECT OF RHEOLOGY Rheology, first proposed by Professor Eugene Cook Bingham of Lafayette College (Bingham, 1922), is the study of the flow and deformation of matters. It is a branch of physics. The term rheology originates from Greek words rheo indicating flow and logia meaning the study of. This definition has been accepted by the American Society of Rheology founded in 1929. In the beginning, rheology was used to describe the properties and behavior of asphalt, lubricants, paints, plastics, rubber, etc. Over the past 100 years, rheology has been developed as an independent and important branch of natural sciences, and is widely applied in many industry sectors and research fields such as: • • • • • • • • • •

Plastics, rubber, and other polymer melts and solutions; Food stuff such as chocolate, ketchup, and yogurt; Metals, alloys, etc.; Concrete, ceramics, glass, and rigid plastics in fresh state and at long periods of loading; Lubricants, greases, and sealants; Personal care stuff and pharmaceuticals; Paints and printing inks; Mud, coal, mineral dispersions, and pulps; Soils, glaciers, and other geological formations; Biological materials such as bones, muscles, and blood.

In a word, rheology has come of age in many ways. Rheology is principally concerned with the following question: “How does a material response to an externally applied force?” That is, rheology describes the relationship between force, deformation, and time. Before introducing the rheology, it is necessary to give a short introduction to the elastic, viscous, and viscoelastic behaviors of solids and/or liquids from the historical perspective, which is summarized in Figure 1.1. An elastic solid subjected to a constant force (within the elastic limit) will undergo a finite deformation, and this deformation will be fully recovered after removal of the applied force. This nature of material is commonly known as elasticity. The force F (N) applied per unit area A (m 2) is defined as stress σ (Pa), and the displacement gradient is called strain γ. In 1678, Robert Hooke established the True Theory of Elasticity (Barnes et al., 1989), stating that “the power of any spring is the same proportion with the tension thereof”, i.e. Hooke’s elastic model. A representative linear elastic solid is spring as presented in Figure 1.2, where the applied force is directly proportional to its distance within the elastic limit. The spring will return its initial shape immediately after removing the applied force. If the applied stress exceeds the elastic limit, the spring will be distorted permanently. The relationship between DOI: 10.1201/9781003265313-1

1

2  Rheology of Fresh Cement-Based Materials

Figure 1.1 Development history of rheology.

Figure 1.2 The Hooke’s elastic model.

the applied stress σ (in Pa) and the strain γ (no unit) for Hooke’s law follows the constitutive equation:

σ = Gγ

(1.1)

where G is the elastic modulus in Pa, which is a measure of the resistance of the solid to deformation. The famous applications of Hooke’s law include retractable pens, manometers,

Introduction to rheology  3

Figure 1.3 The Newton’s viscous law.

spring scales, and the balance wheel of clock. It should be mentioned that Hooke’s law is only applicable to describe the deformation behavior of solids on a small scale. If the strain exceeds the capacity of the material, Hooke’s law will no longer be valid. With respect to an ideal viscous liquid, the material deforms continuously under applied stress, and the deformation cannot be recovered upon removal of the load. In this context, Isaac Newton published a book Philosophie Principia Mathematica in 1687 and proposed Newton’s viscous law, represented by a dashpot as shown in Figure 1.3, to describe this ideal viscous behavior. When applying an external stress to a dashpot, it starts to deform immediately and goes on deforming at a constant velocity v (m/s) until the stress is removed. The rate of the velocity is defined as the shear rate, represented by the symbol γ in s−1, which can be calculated by:

γ = v / h (1.2)

where h is the gap height in m. A higher stress is generally required for a greater strain of the liquid. The coefficient of the applied stress and the shear rate is defined as shear viscosity or dynamic viscosity η in Pa.s, which can be described by:

η = σ / γ (1.3)

The viscosity, associated with the dissipation of kinetic energy of the system, is a quantitative parameter describing the internal fluid friction between elements. Newton’s law is the basic flow model of most of the pure liquids, and these liquids are defined as Newtonian fluids, which will be further discussed later. The stress response to deformation for Hooke’s law and Newton’s law is presented in Figure 1.4. Before the 1900s, Hooke’s law and Newton’s law were widely recognized to describe the behaviors of solids and liquids, respectively. In 1835, Wilhelm Weber found that although the silk thread is a solid-like material, its behavior cannot be ideally described by Hooke’s

4  Rheology of Fresh Cement-Based Materials

Figure 1.4 Stress response to deformation of the Hooke’s law and the Newton’s law.

Figure 1.5 The Maxwell model.

law alone. Indeed, there is no clear distinction between solid and liquid from the viewpoint of rheology. Elastic solid materials can undergo irrecoverable deformations under sufficiently high shear strains, and viscous liquid probably exhibits elastic behavior at extremely low shear strains. This is in agreement with the argument that all materials can flow under sufficient time (Barnes et al., 1989). Since Newton’s law and Hooke’s law are inapplicable to some liquids and solids, the term viscoelasticity was introduced to describe the behavior falling in-between the ideal elastic behavior (Hooke’s law) and the ideal viscous behavior (Newton’s law). The Maxwell model and the Kelvin–Voigt model, as presented in Figures 1.5 and 1.6, respectively, are the most typical models for the viscoelastic materials. The Maxwell model, proposed by James Clerk Maxwell in 1867, consists of a spring and a dashpot connected in series, which is a representation of the simplest viscoelastic liquid model. Under an externally applied stress σ, the Maxwell material shows elastic behavior at very short times and is governed by elastic

Introduction to rheology  5

Figure 1.6  The Kelvin–Voigt model.

modulus G, while after a longer period of time, the viscous behavior becomes predominant and is governed by viscosity η. This means that the strain has two components: one is the immediate elastic component corresponding to the spring, and the other is the viscous component which grows with time. The relationship between the strain and the exerted stress for the Maxwell model can be described by:

dγ σ 1 dσ = + ⋅ (1.4) dt η G dt

If applying a sudden deformation to a Maxwell material, the stress decays on a characteristic timescale of η/G. This is also known as relaxation time, and this phenomenon is defined as stress relaxation. On this basis, a Maxwell material shows elastic solid-like behavior at the deformation time shorter than the relaxation time. The simplest viscoelastic solid can be represented by the Kelvin–Voigt model with a spring and a dashpot connected in parallel, as shown in Figure 1.6. After applying an external stress, the dashpot retards the response of the spring, and thus a delay is required for the strain to develop. That is, the system behaves like a viscous liquid initially (shorter than the relaxation time of η/G) and then elastically over longer time scales. The evolution of strain γ and stress σ with respect to time for the Kelvin–Voigt model can be characterized by:

σ = Gγ + η

dγ (1.5) dt

The Maxwell model applies to liquids with the elastic response, while the Kelvin–Voigt model is used to describe the behavior of solids with a viscous response. However, these two models are not enough to evaluate the viscoelastic behavior of a real system. In this case, some more complicated models such as the Bingham model and the Burger model are proposed to describe the response of viscoelastic fluids. For example, the Bingham model can be regarded as a three-parameter model, which is combined by a parallel unit of a dashpot, a plastic slider, and a spring connected in series (Liingaard et al., 2004), as shown in Figure 1.7. The spring with elastic modulus G is a time-independent component, representing the elastic behavior. The parallel unit of a dashpot with a viscosity η and a slider with a threshold stress σy is a time-dependent component, describing the viscoplastic response. The

6  Rheology of Fresh Cement-Based Materials

Figure 1.7 Conceptual structure of the Bingham model.

Figure 1.8 Stress response to deformation of the Bingham model.

relationship between shear rate and applied stress for the Bingham model can be described as:



 1 dσ  G dt dγ  = dt  1 dσ σ − σ y + η  G dt 

for σ ≤ σ y (1.6) for σ > σ y

It can be seen that the second part of Eq. (1.6) is analogous to the constitutive equation of the Maxwell model in Eq. (1.4). The only difference in the Bingham and Maxwell models is that the stress is replaced by the difference σ − σy. Therefore, the Bingham material shows elastic behavior at σ ≤ σy, while under the applied stress higher than the threshold σy, it exhibits viscous behavior. The stress relaxation to deformation of the Bingham model is shown in Figure 1.8.

Introduction to rheology  7

The flow behavior of pure fluids such as oil and water follows Newton’s law. However, the behaviors of some fluids, e.g. paint, cheese, ketchup, paste, and suspension, do not obey the basic Newtonian flow law. The science of studying the flow behavior of fluid is called rheology. In the following parts, the basic principles of fluid rheology including definitions of viscosity, Newtonian flow, non-Newtonian flow, and the rheology in cement-based materials are briefly illustrated. 1.2 BASIC PRINCIPLES OF RHEOLOGY

1.2.1 Definition of viscosity In general, there are two basic types of flow behavior, i.e. shear flow and extensional flow. This book mainly focuses on the shear flow, where layers of fluid slide one another with different velocities. A typical schematic of shear flow is shown in Figure 1.9. As seen, the upper layer slides at the maximum velocity v (m/s), while the bottom layer remains stationary. The external force F (N) results in a shear stress τ (Pa) in a unit area A (m 2), as depicted in Eq. (1.7). As a response, the upper layer moves a given distance x (m) with the bottom layer remaining stationary. This flow behavior is also defined as laminar flow. If the distance between the upper layer and the bottom layer is denoted as h (m), the deformation gradient across the sample is defined as shear strain γ (dimensionless), and the gradient of the velocity of the material is called shear rate γ in s−1, as calculated by Eqs. (1.8) and (1.9), respectively. The plot of shear stress against shear rate is also called flow curve (Banfill, 1990):

τ = F / A (1.7)



γ = x / h (1.8)



γ = v h (1.9)

Viscosity, indicated by η (Pa.s), is defined as the coefficient ratio between shear stress and shear rate under a state of steady shear, as shown in Eq. (1.10), which is similar to Eq. (1.3). This definition is also called dynamic viscosity:

η = τ / γ (1.10)

Figure 1.9 Typical diagram of shear flow.

8  Rheology of Fresh Cement-Based Materials

Besides, there are a variety of viscosity terms used in practice. For example, the ratio between the dynamic viscosity and the density is defined as the kinematic viscosity:

v=

η (1.11) ρ

where v is the kinematic viscosity in m 2 /s and ρ is the density of the fluid in m3/kg, respectively. The kinematic viscosity is a measure of the internal resistance of a fluid to flow under gravity. The apparent viscosity is the ratio of the shear stress to the shear rate of a certain point on a flow curve. For linear Newtonian flow, which will be discussed later, the value of apparent viscosity is similar to the value of dynamic viscosity. However, for nonlinear flow curves, the apparent viscosity is the slope of a line drawn from the origin to a point on a flow curve, which is dependent on the shear rate. The differential viscosity is defined as the derivative of shear stress against shear rate:

ηdiff =

∂τ (1.12) ∂γ

where ηdiff is the differential viscosity in Pa.s. Plastic viscosity, which is the most popular viscosity parameter in the rheology of cement-based materials, is defined as the limit of differential viscosity with the shear rate approaching infinity:

ηpl = µ = lim

γ →∞

∂τ (1.13) ∂γ

By contrast, the limit of the differential viscosity as the shear rate approaches zero is defined as zero shear viscosity, indicated by η 0:

η0 = lim γ →0

∂τ (1.14) ∂γ

In the case of suspensions, there are also some specific viscosity terms. For example, the ratio of the viscosity of a suspension η to that of the suspending medium ηs is called relative viscosity ηr, described by:

ηr =

η (1.15) ηs

1.2.2 Newtonian flow Fluids can be classified by the concept of viscosity. The Newtonian fluid is the simplest fluid following Newton’s viscous law (see Eq. 1.10), with the viscosity independent of the shear rate or the shear stress. The constant viscosity of a Newtonian fluid is just like the statement in the book Philosophie Principia Mathematica that “the resistance which arises from the lack of slipperiness of the parts of the liquid, other things being equal, is proportional to the velocity with which the parts of the liquid are separated from one another”. The flow curve of Newtonian fluid can be expressed by a series of plots of shear stress versus shear rate, which follows a linear line passing through the origin with the slope of η, as shown in Figure 1.10. However, no real fluid can be described by the Newtonian model perfectly.

Introduction to rheology  9

Figure 1.10 The Newtonian model.

Table 1.1  Viscosity of common substances Substance

Viscosity (mPa.s)

Temperature (°C)

References

Air Water Mercury Benzene Whole milk Coffee cream Olive oil Castor oil Honey Ketchup Bitumen

0.01 1.0016 1.526 0.604 2.12 10 56.2 600 2000–10,000 5000–20,000 10,000,000,000

20 20 25 25 20 25 26 20 20 25 20

Chhabra and Richardson (2011) Rumble (2018) Rumble (2018) Rumble (2018) Fellows (2009) Rumble (2018) Fellows (2009) Chhabra and Richardson (2011) Yanniotis et al. (2006) Koocheki et al. (2009) Chhabra and Richardson (2011)

Instead, some common liquids and gases (e.g. water, alcohol, and air) under ordinary conditions can be assumed to be Newtonian fluid for practical calculations. The viscosity of some common Newtonian fluids is summarized in Table 1.1.

1.2.3 Non-Newtonian flow All fluids that do not follow Newton’s law of viscosity, i.e. constant viscosity independent of shear stress, are defined as non-Newtonian fluid. In this kind of fluid, the viscosity changes with the applied shear stress or shear rate history. Indeed, compared to the Newtonian flow curve passing through the origin with a constant scope, the flow curve of non-Newtonian fluids is different. For some non-Newtonian fluids with shear-independent viscosity, e.g. Bingham flow, the flow curve still exhibits a different behavior compared to Newtonian flow. In this case, the concept of viscosity is inadequate to describe the rheological behavior of non-Newtonian fluids. In this context, various models or constitutive equations have been developed to distinguish the behavior of non-Newtonian fluids and idealize the flow curves, as presented in Figure 1.11. As mentioned earlier, all non-Newtonian fluids exhibit a distinct flow behavior compared with the Newtonian flow. In reality, some fluids possess yield stress, which is the minimum stress when flow occurs. In other words, applying an external shear stress lower than the yield stress, the fluid cannot flow and only deform elastically. When the applied shear stress

10  Rheology of Fresh Cement-Based Materials

Figure 1.11  Typical common constitutive models for non-Newtonian fluids.

exceeds the yield stress, the internal network structure will be destroyed and the material starts to flow like a fluid. Materials with yield stress are also considered to be visco-plastic materials. For the material with zero yield stress but inconstant (or shear rate-dependent) viscosity, its rheological behavior can be characterized by the power law model with an exponential relationship between shear stress and shear rate (Ostwald, 1929), as expressed by:

τ = aγ b (1.16)

where a is the consistency index in Pa.sb and b is the flow index (dimensionless). As shown in Figure 1.11, if the exponent b is higher than 1, the flow curve is concave upward, and the viscosity increases with the shear rate, indicating that the material shows shear-thickening or dilatant behavior. By contrast, if the exponent b is lower than 1, the flow curve is concave downward. It means that the material exhibits shear-thinning or pseudo-plastic behavior, i.e. the viscosity decreases with the increase of shear rate. For suspensions, the shearinduced behavior is related to the distribution of particles in the suspending medium, with particle rearrangement along the direction of applied shear for shear-thinning and clusters/­ jamming formation for shear thickening, as shown in Figure 1.12. The shear-induced behavior depends on the concentration and intrinsic properties of the dispersed particles such as shape, size, and density. Generally, a suspension with a particle concentration higher than 75% tends to show shear-thickening behavior. For a material with a yield stress and a shear rate (stress) independent viscosity, its rheological behavior can be described by the Bingham model (Bingham, 1917). The constitutive equation is expressed as:

τ = τ 0 + µγ (1.17)

where τ0 and μ are the yield stress in Pa and the plastic viscosity in Pa.s, respectively. The plastic viscosity in the Bingham model refers to the same physical meaning as the viscosity in a Newtonian model, i.e. the slope of the curve between shear stress and shear rate. Toothpaste is a classic example of Bingham fluid.

Introduction to rheology  11

Figure 1.12  Shear-thinning and shear-thickening behaviors of suspensions.

For material with a yield stress and a shear rate-dependent viscosity, the flow behavior can be characterized by the Herschel–Bulkley model (Herschel and Bulkley, 1926), as expressed by:

τ = τ 0 + aγ b (1.18)

where a is the consistency index in Pa.sb and b is the flow index (dimensionless). The material exhibits shear-thinning behavior when b  1. If b = 1, this model reduces to the Bingham equation. It should be mentioned that the plastic viscosity cannot be directly obtained from the Herschel–Bulkley model. Instead, an equivalent plastic viscosity μe can be calculated from the consistency index and the flow index as follows:

µ=

3a b−1 γ max (1.19) b+2

where γ max is the maximum shear rate. Another model describing the nonlinearity of the flow of fluids with yield stress is the Casson model:

τ 1/2 = τ 01/2 + µ1/2γ 1/2 (1.20)

In addition to the above-stated models, Table 1.2 summarizes some other equations describing the rheological behavior of non-Newtonian fluids.

1.2.4 Thixotropy The viscosity of the pseudo-plastic fluids is time-independent. However, certain fluids are showing time-dependent shear-thinning behavior. This kind of fluid is called thixotropic fluid. The concept of thixotropy was first introduced by Freundlich (Freundlich and Juliusburger, 1935), from two ancient Greek words: thixis meaning touch and tropo meaning

12  Rheology of Fresh Cement-Based Materials Table 1.2  Summary of rheological equations for non-Newtonian fluids Model

Equation

References

τ = Aγ τ = τ 0 + µγ τ = τ 0 + aγ + bγ 2

Ostwald (1929)

Bingham Modified Bingham Herschel–Bulkley

τ = τ 0 + aγ b

Herschel and Bulkley (1926)

Casson

τ

Power

n

1/2



1/2 0

Bingham (1917) Feys et al. (2007), Feys et al. (2008) Papo and Piani (2004)

+ µ γ

1/2 1/2

Generalized Casson

m

τ m = τ 0 m + [η∞γ ]

Papo and Piani (2004)

Papo–Piani

τ = τ 0 + η∞γ + Kγ n τ = τ 0 + ηγ e −αγ

Papo and Piani (2004)

De Kee

Yahia and Khayat (2003)

τ = τ 0 + B sinh (γ / C ) τ = aγ + B sinh−1(γ / C )

Vom Berg (1979)

Roberston–Stiff

τ = a (γ + C )

Yahia and Khayat (2003)

Atzeni

Atzeni et al. (1983)

Williamson

γ = aτ 2 + bτ + δ γ τ = η∞γ + τ f γ + Γ

Sisko–Ellis

τ = aγ + bγ c ( c < 1)

Papo (1988)

Vom Berg Eyring

−1

b

Atzeni et al. (1985)

Papo (1988)

η = η∞ + Kγ n−1 Shangraw–Grim–Mattocks

τ = τ 0 + η∞γ + α1 1− exp ( −α 2γ ) 

Yahia–Khayat

τ = τ0 + 2

(

τ 0η∞

)

γ e −αγ

Papo (1988) Yahia and Khayat (2003)

to change. Simply, thixotropy means that change occurs by touch. Technically, thixotropy is defined as the continuous decrease in viscosity with time when applying an external shear stress to a rest suspension, and then a subsequent recovery of viscosity over time when the shear stress is removed (Mewis and Wagner, 2009), as illustrated in Figure 1.13. Two characteristics can be recognized from this definition. First, the thixotropic material shows a time-dependent decrease of viscosity under flow state, i.e. time-dependent shear-thinning viscosity. Second, thixotropy is reversible. Thixotropy only appears in shear-thinning fluids, due to the fact that the breakdown structure cannot reform immediately after the removal of the shear stress. Daily-used products such as toothpaste, paints, ketchup, and hair gel are thixotropic fluids. Thixotropy can be evaluated by hysteresis loop area (Jiao et al., 2019b, Roussel, 2006), structural breakdown area (Khayat et al., 2012), static yield stress (Lowke, 2018), thixotropic index (Patton, 1964), stress recovery (Qian and Kawashima, 2016), storage modulus (Jiao et al., 2019a, Yuan et al., 2017), etc. Readers are referred to Jiao et al. (2021) and Mewis and Wagner (2009) for further details. Thixotropy is important for applications that require the material to flow easily during placing, but then the structure is built quickly when the flow stops. A typical usage of thixotropy in engineering is the extrusion-based 3D printing process (De Schutter et al., 2018, Yuan et al., 2019), where the material needs to flow easily in the pump line and nozzle and builds up its mechanical strength immediately after arriving at the final position.

Introduction to rheology  13

Figure 1.13  Schematic diagram of thixotropy. (Adapted from Lowke, 2018.)

Figure 1.14  Thixotropy versus anti-thixotropy.

1.2.5 Anti-thixotropy (rheopexy) Opposite to thixotropy, the time-dependent shear-thickening behavior is defined as rheopexy, also called anti-thixotropy or negative thixotropy. The time-dependent increase in viscosity during flow is generally caused by flow-induced aggregation. The difference between thixotropy and anti-thixotropy behaviors is presented in Figure 1.14. The anti-thixotropic materials are much less common. Typical examples of rheopectic fluids include gypsum pastes, cream, and printer inks, which become more stiffening by continuous shaking or mixing. Besides, suspensions with a solid volume fraction lower than 10% generally exhibit a negative thixotropic behavior. Overall, a schematic representation of the flow behavior of various fluids is provided in Figure 1.15. 1.3 CEMENT-BASED MATERIALS

1.3.1 History of cement and concrete Joseph Aspdin, an English bricklayer and builder, is generally recognized as the first inventor of Portland cement. In 1824, he patented the method of binder production from the burned

14  Rheology of Fresh Cement-Based Materials

Figure 1.15 Schematic representation of flow behavior of various fluids.

Figure 1.16 (a and b) Joseph Aspdin and a plaque in Leeds.

mixture of limestone and clay, and this powder can react with water to develop strength. This product is named Portland cement for the first time because its color resembled that of stone from the location of Portland. Of course, the cement invented by Aspdin was nothing more than a hydraulic lime, and its mineralogy was completely different from that of the toady’s cement. Even so, Aspdin’s patent gave him the priority for using the term Portland cement. Aspdin’s cement patent is undoubtedly a pioneering work that opens a door for the development of modern cement. Consequently, Joseph Aspdin was acclaimed as the father of Portland cement, as shown in Figure 1.16. The first cement plant was built in Grodziec, Poland and started its production in 1857. However, the cement quality at that time was very low. Without the great progress in its production, it cannot be used for large-scale applications (Hewlett and Liska, 2019). In the history of Portland cement, the production technique and quality control are the most important to the large-scale production of modern cement. The invention of the rotary kiln aided in the large-scale production of modern cement possible (Hewlett and Liska,

Introduction to rheology  15

Figure 1.17  Cement production worldwide from 1995 to 2020 (in billion tons).

2019). Cement manufacture was changed from a batch process to a continuous production process, and the quality of cement was significantly improved. In 1877, Thomas Crampton first patented a kiln with a revolving cylinder lined with firebricks. In 1885, Frederick Ransome, an American, produced a 6.5 m long cylindrical kiln, but it didn’t work well. In 1898, Hurry and Seaman, both Americans, successfully built the first fully operational rotary kiln supplied with energy by coal. From then on, cement has become the major construction material, and the cement industry has boomed to meet the huge demands for the construction of buildings, roads, bridges, dams, and factories all over the world. Figure 1.17 gives the worldwide output of cement from 1995 to 2020 (US Geological Survey). In 2019, China yielded 56.2% of the worldwide output of cement (CEMBUREAU), as shown in Figure 1.18, due to the massive construction of infrastructure recently. Considering both economic and technical reasons, cement is mainly used together with sand and stone to form mortar and concrete. On the one hand, sand and stone are much cheaper than cement, and thus the cost can be reduced. On the other hand, sand and stone are more stable in terms of volume or chemistry, and their application can improve the volume and chemical stability of mortar or concrete. Mortar is a mixture of sand and a stiff paste of cement made with a proper amount of water, which is often used to bind bricks or blocks. Mortar also has some special applications such as grouting mortar and floor mortar. Concrete is a mixture of coarse aggregate (e.g. small stones or gravel), fine aggregate (e.g. sand), cement, and enough water, which is often used for the construction of primary structural members. The vast need of infrastructure and the various requirements for construction materials have stimulated the invention and development of many benchmark concrete techniques. These technological advancements make concrete the most important and widely used construction material. The most important dates in the history of the development of contemporary concrete are the following: 1867—Joseph Monier introduced the reinforcement of concrete, which is the major composite used in infrastructure. 1870—The production of precast elements began. Precast elements are still very popular currently, since concrete elements can be manufactured in the factory. The structural

16  Rheology of Fresh Cement-Based Materials

Figure 1.18 World cement production in 2019 by region and main countries.

quality, noise, and pollution can be better controlled. Many governments and authorities are trying to increase the percentage of the precast element and reduce the cast-in-site element. 1907—Koenen invented the prestressed concrete. This form of element fully uses the compressive strength of concrete and the tensile strength of steel. Prestressed concretes are used in many heavy loading structures, such as large-span bridges and beams. 1924—Bolomey proposed the formula for concrete strength calculation. The Bolomey equation is still one of the most important equations used for modern concrete 100 years later. 1950—Water-reducing agent, which was polymerized by naphthalene formaldehyde sulfonate salts, was first used to increase the flowability of concrete. The technology for chemical admixtures developed very fast, with water-reducing agents evolving into the superplasticizer. The application of superplasticizer makes high-strength concrete possible. 1980—Aïtcin first introduced high-performance concrete (HPC). HPC exceeds the properties and constructability of normal concrete. Normal and special materials are used to make these specially designed concretes that must meet a combination of performance requirements. 1988—Okamura invented self-compacting concrete (SCC). This type of concrete can flow and fill up the formwork under its own weight without vibration. Due to its economic and technical benefits, SCC has been widely used in industries, such as precast industry and heavily reinforced section. 1994—De Larrard first introduced ultra-high-performance concrete (UHPC) with low porosity, high compressive strength (above 120  MPa), high durability, and self-­ compactability, based on an optimized particle-packing model (de Larrard and Sedran, 1994). In 1997, UHPC was first used to construct a pedestrian bridge in Canada.

1.3.2 Fresh properties of cement-based materials Fresh properties of concrete determine whether strength grade, elastic modulus, durability, and even color of hardened concrete meet the design or not. In the early history of concrete, workers assessed the fresh properties of concrete by subjective methods, which made quantitative assessment hard. Abrams was probably the first well-known researcher who

Introduction to rheology  17

recognized the importance of the workability of fresh concrete. He proposed the slump test to quantitatively evaluate the workability of fresh concrete. The slump test is still used up to now and universally accepted as a quality control measure of fresh concrete owing to the following reasons: • The testing setup for the slump test is quite cheap and simple. • The slump test is quite easy to perform. • The test result, to some extent, not only reflects the flowability of concrete, but also roughly judges the bleeding and cohesiveness of concrete. As pointed out by Tattersall (Tattersall and Banfill, 1983), although the slump test has been the most widely used method for the workability of concrete, it is an empirical test without a scientific base and the testing result is barely related to basic physical constants. Different operators may give different slump values for the same concrete. Moreover, two concretes with the same slump value may work quite differently. Therefore, rheology is introduced to describe the fresh properties of concrete from other disciplines for reference. Through a large number of investigations, it was found that many mature analytical and physical models in rheology can be applied to concrete, and various rheological testing techniques can also be employed to characterize the viscoelastic behavior of concrete. Common rheometers can be directly applied to cement paste which is a suspension with micro-sized particles suspended in water. Therefore, the rheological behavior of cement paste was studied as early as the 1950s. Tattersall published his pioneering work on the structural breakdown of cement paste in the journal of Nature (Tattersall, 1955). In this work, a theoretical understanding of the structural breakdown of cement paste was attempted. Afterward, many other papers on the rheological behavior of cement paste were published. Due to the large size of coarse aggregate in fresh concrete, common rheometers are not suitable for measuring concrete. By the 1970s, a rotating vane or coaxial cylinder with a wide gap was invented for concrete and used to measure the relationship between torque and rotational speed, whereas the complex mathematic transformation was not made for the testing results. At that time, researchers were just trying to find a more accurate testing method to replace the slump test. Tattersall successfully developed a practical testing setup, which was called the two-point test (Figure 1.19), that could be employed both in the lab

Figure 1.19  Schematic diagram of Tattersall’s two-point test rheometer. (Adapted from Ferraris, 1999.)

18  Rheology of Fresh Cement-Based Materials

and on-site (Tattersall and Bloomer, 1979). This test measures the shear stress under at least two shear rates, allowing to calculate the rheological parameters, e.g. yield stress and plastic viscosity. This testing setup started to bring the concept of rheology to concrete science and technology. However, due to the low content of paste, concrete still cannot flow like water. Most of the aggregates in concrete directly contact with each other when concrete flows. Thus, concrete flows with a friction and collision regime. Rheology based on continuum mechanics is not a good option for characterizing the flow behavior of this material. Tattersall argued the advantages of two-point testing of fresh concrete workability over the empirical test methods that had been employed to date (Tattersall and Banfill, 1983). The invention of SCC in 1988 immediately buildup a close link between concrete and rheology. SCC is a type of concrete that is rich in cement paste. Most of the aggregates are suspended in the paste. Therefore, SCC can flow like water just under the self-weight, and the aggregates are no longer in direct contact during the flow process. Therefore, the rheological theory and experiment are perfectly applicable to SCC. The enthusiasm for rheology research was inspired by the invention and wide application of SCC. The rheology of SCC is completely different from ordinary concrete, bringing totally a different engineering behavior. High flowability of SCC is required to ensure self-compacting property. The density of water is 1000 kg/m3, while it is 2600–2700 and 3150 kg/m3 for aggregate and cement particles, respectively. On the other hand, bleeding and segregation stemming from the large density difference of constituents more possibly happen for SCC than ordinary concrete if the proper mix design of SCC is not taken. The high viscosity of the paste is required by SCC to improve the resistance of bleeding and segregation. Followed by Tattersall’s two-point rheometer, many other more sophisticated testing setups with a more accurate analytical model, computer-controlled systems, and automatic data collection were developed. For example, Wallevik developed a new coaxial cylinder rheometer for fresh concrete in the 1990s (Wallevik and Gjørv, 1990). This rheometer has been successfully commercialized as the BML, and now it upgrades to its higher version, ConTec Visco 5, as shown in Figure 1.20a. A typical dimension of the rheometer has a bob diameter of 100 mm and a container with a diameter of 145 mm, resulting in a capacity of 17 L of concrete. It is designed for concrete with slumps greater than 120 mm. In this rheometer, the outer cylinder rotates, while the inner bob registering the torque remains stationary. In 1993, the biggest coaxial rheometer with a container radius of 1200 mm, a bob of 760 mm, and a capacity of 500 L of concrete was developed in France. It was named CEMAGREF (Coussot, 1993), as presented in Figure 1.20b. However, this rheometer has not been commercially available due to its large volume. In 1994, the IBB rheometer (see Figure 1.20c) was developed by Beaupre in Canada (Beaupre, 1994). The vane motion of IBB was planetary and not axial. The testing sample was about 21 L of concrete. It was designed for concrete with slumps between 20 and 300 mm. In 1996, the BTRHEOM was developed under the direction of de Larrard et al. at the Laboratoire des Ponts et Chaussées (LCPC) in Paris (Hu et al., 1996). The BTRHEOM rheometer, as shown in Figure 1.20d, was the only concrete rheometer based on parallel plate geometry. This rheometer required about 7 L of concrete that has at least 100 mm of the slump, and it was designed to be portable so that it could be used at a construction site. In 2004, the ICAR rheometer, as depicted in Figure 1.20e, was developed at the International Center for Aggregates Research (ICAR) at the University of Texas at Austin by Koehler and his colleagues (Koehler and Fowler, 2004). This rheometer was well designed and portable for on-site testing. Different rotators and containers are available for concrete with different aggregate sizes. The ICAR rheometer is available for concrete with a slump higher than 75 mm.

Introduction to rheology  19

Figure 1.20 Concrete rheometers: (a) ConTec Visco 5, (b) CEMAGREF (adapted from Coussot, 1993), (c) IBB rheometer, (d) BTRHEOM rheometer, and (e) ICAR rheometer.

Since the behavior of concrete may range from a very stiff state to SCC and various test rheometers are used, the rheological results from different authors are hard to compare. Thus, organizations from different countries carried out several programs for the c­ omparison of different rheometers. These test programs have been executed with several rounds. The first round-robin test was carried out in 2000 in Nante, France (Ferraris and Brower, 2000). It was found that different types of mixtures were ranked statistically in the same order by all the rheometers for both yield stress and plastic viscosity. However, obvious differences in absolute values given by the various rheometers were found. Another round-robin test carried out in 2003 in Cleveland OH, USA (Ferraris and Brower, 2003), confirmed the findings obtained in 2000. It was also pointed out that small variations in the concrete could cause large changes in the rheometer results. The RILEM TC MRP (Measuring Rheological Properties of Cement-based materials) organized a new round-robin test in Bethune, France in 2018 to evaluate the rheological properties of different mortar and concrete mixtures (Feys et al., 2019). The rheological properties of fresh concrete evolve with time, due to the continuous hydration of cementitious materials. During this period, hydration products fill in the voids and connect in-between the particles (Huang et al., 2020a, Huang et al., 2020b), and a percolated network is progressively formed. It has been shown that the C–S–H bridge is the main reason for the rigidification of cement paste at the early age (Mostafa and Yahia, 2017, Roussel et al., 2012). Thus, the viscoelasticity and rheological properties of fresh cement paste change with continuous cement hydration, which is important to many engineering application scenarios of cement-based materials, e.g. mixing, pumping, casting, 3D concrete printing, and smart casting (Jiao et al., 2021, Roussel, 2006, Roussel, 2018). It is well known that concrete is strong in compression and weak in tension. With the increase of compressive strength, concrete becomes more brittle, leading to catastrophic

20  Rheology of Fresh Cement-Based Materials

failures. Fiber-reinforced concrete (FRC) has been studied for many years to overcome this tension weakness of all types of concrete. Fiber-reinforced concrete is a promising composite, whereas the incorporation of fibers in concrete changes the rheological properties. The suspension of FRC is a non-Newtonian fluid, which can display differences in normal stress. Flow resistance arises due to the opposite movement of the particles. Thus, the rheology of fiber-reinforced concrete is also different from that of ordinary concrete. It can be said that the rheology of cement-based materials is far more complicated than other materials for the following reasons: • Complex constituents. Cement-based materials are inorganic ones which may include various types of particles such as cement, fly ash, slag, silica fume, aggregates, and fibers. Furthermore, many organic polymers are often used as their key admixtures. • Evolving with time. Hydration of cement begins and continuously goes on after the contact of cement and water. The ongoing hydration results in the time-varying rheology of cement-based materials. • Multi-scale particles ranging from nm to mm. Cement-based materials include microsized particles such as cement and fly ash, and sand and coarse aggregate with the size of mm. However, the complexity of the rheology of cement-based materials has not dampen the researchers’ enthusiasm. Numerous researches have been carried out, and are carried out, on the rheology of cement-based materials, including the fundamentals, measurements, and predictions of rheology. Rheology for cement-based materials is not just “a study of flow and deformation”. In addition, the underlying physical and chemical reactions behind the flow and deformation become more and more crucial for the study of rheology. This brings the rheology to the molecular level, i.e. micro-rheology. It means that the key interest is devoted not only to the movements of physical points but also to the chemical and physical reactions that happen inside a point during the deformation of the medium. The developed techniques and new findings in the field of rheology play an important role in the advancement of new concrete technologies. 1.4 THE SCOPE OF THIS BOOK As stated above, rheology is a very useful technical tool and scientific basis for cementbased materials. In particular, new technologies put forward new requirements for rheological properties. The complicated and interesting rheology of cement-based materials attracts the attention of more and more researchers, and much funding around the world has been allocated to this topic. This book focuses on the flow and rheological behavior of fresh cement-based materials. It should be mentioned that the rheology of hardened cement and concrete is beyond the scope of this book, and hence will not be discussed. Three most important parts of the rheology of cement-based materials are discussed in this book. The first part deals with the fundamentals of the rheology of cement-based materials. The basic knowledge of rheology is briefly introduced in this part. Cement-based materials are classified into two main groups based on the particle sizes, i.e. cement paste and concrete (including mortar). Cement paste is a suspension with an interstitial solution and micro-sized particles. Colloidal interaction between particles has a great influence on the rheology of cement paste. On a larger scale, sand and aggregate are in the size of mm. Lubrication, friction, and collision may dominate the flow regime of concrete, depending on

Introduction to rheology  21

the rheological properties of interstitial fluid (i.e. cement paste) and the aggregate ratio. All these are discussed in detail in this part. The second part discusses the measurement techniques for the rheology of cement-based materials. More than 100 techniques have been developed to evaluate the rheological properties of cement-based materials, either rheometers with sound scientific bases or empirical methods. In addition, due to the large size of aggregate in concrete, common rheometers do not apply to concrete. Therefore, many rheometers specialized in concrete have been developed. Obtaining the accurate rheological parameters of cement-based materials is crucial to this field. The third part deals with the application of rheology in concrete technology. New technologies always come with new requirements for materials, even new materials. From SCC, new cementitious materials, to digital fabrication, they put forward new requirements on the rheological properties of concrete.

REFERENCES Atzeni, C., et al. (1983). “New rheological model for Portland cement pastes.” Il Cemento, 80. Atzeni, C., et al. (1985). “Comparison between rheological models for Portland cement pastes.” Cement and Concrete Research, 15(3), 511–519. Banfill, P. (1990). “The rheology of cement paste: Progress since 1973”, in Properties of Fresh Concrete: Proceedings of the International RILEM Colloquium, edited by HJ Wierig, Chapman & Hall. CRC Press, Boca Raton, FL, pp. 3–9. Barnes, H. A., et al. (1989). An introduction to rheology. Elsevier, Amsterdam, the Netherlands. Beaupre, D. (1994). Rheology of high performance shotcrete. Doctoral dissertation, University of British Columbia, Vancouver, Canada. Bingham, E. C. (1917). An investigation of the laws of plastic flow. US Government Printing Office, Washington, United States. Bingham, E. C. (1922). Fluidity and plasticity. McGraw-Hill, New York, United States. Chhabra, R. P., and Richardson, J. F. (2011). Non-Newtonian flow and applied rheology: Engineering applications. Butterworth-Heinemann, Oxford, United Kingdom. Coussot, P. (1993). Rhéologie des boues et laves torrentielles: étude de dispersions et suspensions concentrées. Editions Quae, Versailles, France. de Larrard, F., and Sedran, T. (1994). “Optimization of ultra-high-performance concrete by the use of a packing model.” Cement and Concrete Research, 24(6), 997–1009. De Schutter, G., et al. (2018). “Vision of 3D printing with concrete - Technical, economic and environmental potentials.” Cement and Concrete Research, 112, 25–36. Fellows, P. J. (2009). Food processing technology: Principles and practice. Elsevier, Amsterdam, the Netherlands. Ferraris, C. F. (1999). “Measurement of the rheological properties of high performance concrete: State of the art report.” Journal of Research of the National Institute of Standards and Technology, 104(5), 461–478. Ferraris, C. F., and Brower, L. E. (2000). Comparison of Concrete Rheometers: International test at LCPC (Nantes France) in October, 2000. Ferraris, C. F., and Brower, L. E. (2003). Comparison of Concrete Rheometers: International tests at MB (Cleceland OH, USA) in May, 2003. Feys, D., et al. (2007). “Evaluation of time independent rheological models applicable to fresh selfcompacting concrete.” Applied Rheology, 17(5), 56244–56241–56244–56210. Feys, D., et al. (2008). “Fresh self compacting concrete, a shear thickening material.” Cement and Concrete Research, 38(7), 920–929.

22  Rheology of Fresh Cement-Based Materials Feys, D., et al. (2019). “An overview of RILEM TC MRP Round–Robin testing of concrete and mortar rheology in Bethune, France, May 2018.” Proceedings 2nd International RILEM Conference Rheology and Processing of Construction Materials (RheoCon2), Dresden, Germany. Freundlich, H., and Juliusburger, F. (1935). “Thixotropy, influenced by the orientation of anisometric particles in sols and suspensions.” Transactions of the Faraday Society, 31, 920–921. Herschel, W. H., and Bulkley, R. (1926). “Konsistenzmessungen von gummi-benzollösungen.” Kolloid-Zeitschrift, 39(4), 291–300. Hewlett, P., and Liska, M. (2019). Lea’s chemistry of cement and concrete. Butterworth-Heinemann, Oxford, United Kingdom. Hu, C., et al. (1996). “Validation of BTRHEOM, the new rheometer for soft-to-fluid concrete.” Materials and Structures, 29(10), 620–631. Huang, T., et al. (2020a). “Understanding the mechanisms behind the time-dependent viscoelasticity of fresh C3A–gypsum paste.” Cement and Concrete Research, 133, 106084. Huang, T., et al. (2020b). “Evolution of elastic behavior of alite paste at early hydration stages.” Journal of the American Ceramic Society, 103(11), 6490–6504. Jiao, D., et al. (2019a). “Structural build-up of cementitious paste with nano-Fe3O4 under timevarying magnetic fields.” Cement and Concrete Research, 124, 105857. Jiao, D., et al. (2019b). “Time-dependent rheological behavior of cementitious paste under continuous shear mixing.” Construction and Building Materials, 226, 591–600. Jiao, D., et al. (2021). “Thixotropic structural build-up of cement-based materials: A state-of-the-art review.” Cement and Concrete Composites, 122, 104152. Khayat, K. H., et al. (2012). “Evaluation of thixotropy of self-consolidating concrete and influence on concrete performance.” Proceedings of the 3rd Iberian Congress on Self Compacting Concrete, Madrid, Spain, 3–16. Koehler, E. P., and Fowler, D. W. (2004). “Development of a portable rheometer for fresh portland cement concrete.” Technical Report, International Center for Aggregates Research (ICAR), The University of Texas at Austin, Austin, United States. Koocheki, A., et al. (2009). “The rheological properties of ketchup as a function of different hydrocolloids and temperature.” International Journal of Food Science & Technology, 44(3), 596–602. Liingaard, M., et al. (2004). “Characterization of models for time-dependent behavior of soils.” International Journal of Geomechanics, 4(3), 157–177. Lowke, D. (2018). “Thixotropy of SCC—A model describing the effect of particle packing and superplasticizer adsorption on thixotropic structural build-up of the mortar phase based on interparticle interactions.” Cement and Concrete Research, 104, 94–104. Mewis, J., and Wagner, N. J. (2009). “Thixotropy.” Advances in Colloid and Interface Science, 147–148, 214–227. Mostafa, A. M., and Yahia, A. (2017). “Physico-chemical kinetics of structural build-up of neat cement-based suspensions.” Cement and Concrete Research, 97, 11–27. Ostwald, W. (1929). “de Waele-Ostwald equation.” Kolloid Zeitschrift, 47(2), 176–187. Papo, A. (1988). “Rheological models for cement pastes.” Materials and Structures, 21(1), 41–46. Papo, A., and Piani, L. (2004). “Flow behavior of fresh Portland cement pastes.” Particulate Science and Technology, 22(2), 201–212. Patton, T. C. (1964). “Paint flow and pigment dispersion.” Paint Flow and Pigment Dispersion, pp. 479–479. Qian, Y., and Kawashima, S. (2016). “Use of creep recovery protocol to measure static yield stress and structural rebuilding of fresh cement pastes.” Cement and Concrete Research, 90, 73–79. Roussel, N. (2006). “A thixotropy model for fresh fluid concretes: Theory, validation and applications.” Cement and Concrete Research, 36(10), 1797–1806. Roussel, N., et al. (2012). “The origins of thixotropy of fresh cement pastes.” Cement and Concrete Research, 42(1), 148–157. Roussel, N. (2018). “Rheological requirements for printable concretes.” Cement and Concrete Research, 112, 76–85. Rumble, J. R. (2018). CRC handbook of chemistry and physics. CRC Press, Boca Raton, FL.

Introduction to rheology  23 Tattersall, G. (1955). “Structural breakdown of cement pastes at constant rate of shear.” Nature, 175(4447), 166–166. Tattersall, G. H., and Banfill, P. F. (1983). The rheology of fresh concrete. Pitman Books Limited, London, England. Tattersall, G. H., and Bloomer, S. (1979). “Further development of the two-point test for workability and extension of its range.” Magazine of Concrete Research, 31(109), 202–210. Vom Berg, W. (1979). “Influence of specific surface and concentration of solids upon the flow behaviour of cement pastes.” Magazine of concrete research, 31(109), 211–216. Wallevik, O., and Gjørv, O. (1990). “Development of a coaxial cylinders viscometer for fresh concrete.” Proceedings of the Properties of Fresh Concrete: Proceedings of the International RILEM Colloquium, Edited by H.J. Wierig, CRC Press, Hanover, Germany, pp. 213–224. Yahia, A., and Khayat, K. H. (2003). “Applicability of rheological models to high-performance grouts containing supplementary cementitious materials and viscosity enhancing admixture.” Materials and Structures, 36(260), 402–412. Yanniotis, S., et al. (2006). “Effect of moisture content on the viscosity of honey at different temperatures.” Journal of Food Engineering, 72(4), 372–377. Yuan, Q., et al. (2017). “On the measurement of evolution of structural build-up of cement paste with time by static yield stress test vs. small amplitude oscillatory shear test.” Cement and Concrete Research, 99, 183–189. Yuan, Q., et al. (2019). “A feasible method for measuring the buildability of fresh 3D printing mortar.” Construction and Building Materials, 227, 116600.

Chapter 2

Rheology for cement paste

2.1 INTERACTION BETWEEN PARTICLES IN THE PASTE Cement paste is a suspension system containing water and cement particles, and even mineral admixture particles in some cases. Its rheological properties have universal characteristics as suspensions. The volume fraction of suspending particles, the particle shape and size distribution, and the rheological properties of the liquid medium can directly influence the rheological properties of the cement paste. In addition, hydration of cementitious particles and functions of chemical admixtures lead the rheological properties of cement paste to be more complex. The interparticle interactions can be changed due to the presence of hydrates and chemical admixtures. From a micromechanical view, the rheological properties of cement paste are determined by the interparticle interactions, including colloidal interactions, Brownian forces, hydrodynamic forces, and contact forces.

2.1.1 Colloidal interaction Several types of non-contact interactions occur within a cementitious suspension (Flatt, 2004b). At a short distance, cement particles interact via (generally attractive) van der Waals forces (Flatt, 2004a). Also, there are electrostatic forces that result from the absorption of ions on the particle surfaces (Flatt and Bowen, 2003). Polymer additives used in many modern cementitious materials can introduce steric hindrance (Banfill, 1979, Yoshioka et al., 1997), which is believed to predominate over electrostatic repulsion. Each of these different interactions introduces non-contact forces between particles, the magnitude of which depends primarily on their separation distance. Van der Waals interactions were shown to dominate all other colloidal interactions in the case of cement pastes and therefore dictate the interparticle distance (Flatt, 2004a, Flatt et al., 2009). The interparticle force is given by:

F≅

A0 a∗ (2.1) 12H 2

where a* is the radius of curvature of the “contact” points, H is the surface-to-surface separation distance at “contact” points, and A 0 is the non-retarded Hamaker constant (Roussel et al., 2010). Without inclusion of polymers, the value of H is estimated to be in the order of a couple of nm (Flatt, 2004b, Roussel et al., 2010). However, the steric hindrance induced by the absorption of polymers affects the value of H, and H is dictated by the conformation of the adsorbed polymer on the surface of the cement grains (Flatt et al., 2009). Orders of magnitude for typical polymers are around 5 nm (Flatt et al., 2009). Recent advances have shown DOI: 10.1201/9781003265313-2

25

26  Rheology of Fresh Cement-Based Materials

that it shall be possible to correlate the molecular structure of the polymer to its surface conformation (Flatt et al., 2009). 2.1.1.1 Van der Waals force The van der Waals force is consisted of three components (Hiemenz and Rajagopalan, 1997, Wallevik, 2003, Jiao et al., 2021a), i.e. Keesom, Debye, and London interactions, of which the London dispersion force is the most important one. This force is caused by the interaction between induced dipoles in the neighboring molecules, and it always leads to an attraction between particles. The van der Waals potential is the product of the Hamaker constant and the geometrical factor. The Hamaker constant depends on the dielectric properties of the particles and the suspending medium, while the geometrical factor depends on the size and shape of the particles and the separation distance between the particles. For cement, due to the large complexity of the material, no values for the Hamaker constant have been measured (Flatt, 2004a). Published results are based on assumptions and measurements of other minerals. 2.1.1.2 Electrostatic repulsion Particles in suspensions are electrically charged on their surface. In the suspending medium, counter-ions, which are ions with opposite charges, are attracted to the particles, resulting in a decreasing concentration of these counter-ions with increasing distance from the particle, creating the diffuse double layer (Hiemenz and Rajagopalan, 1997). When two particles approach, their diffuse double layers will start to overlap, creating a higher concentration of counter-ions. As a result, osmotic pressure is created to neutralize this overconcentration of counter-ions and the particles are repelled from each other. 2.1.1.3 Steric hinder force Adding a polymer layer to a particle surface can create a geometrical barrier for other particles. As a result, two particles cannot approach closer than the separation distance created by the polymers. In contrast to the van der Waals forces and the electrostatic repulsion forces, the absolute value of the potential caused by the polymers is not a monotone decreasing function of the interparticle distance. Instead, it shows a more abrupt change from its characteristic value near the particle surface, to zero at a particle distance equal to the effective length of the polymers (Hiemenz and Rajagopalan, 1997). This repulsion mechanism is called steric hindrance and is a purely geometrical phenomenon.

2.1.2 Brownian forces Every material is subjected to thermal agitation when the temperature is higher than 0 K. The thermal energy equals kT, where k is the Boltzmann constant and T is the temperature in Kelvin. For larger particles suspended in a certain medium, the energy exerted by temperature is relatively low, compared to gravity and viscous drag forces. The thermal activation does not influence the behavior of the suspension. However, for particles smaller than 1 µm, the Brownian force (kT/a, which equals Brownian/thermal energy, divided by the particle radius a) has a similar or higher order of magnitude as gravity, meaning that it can no longer be neglected. A dimensionless parameter Nr estimating the relative magnitude of Brownian motion over interparticle force is expressed by Eq. (2.2):

Nr =

A0 d (2.2) 12H ⋅ kT

Rheology for cement paste  27

where A 0 is the Hamaker constant (J), d is the particle diameter (m), k is the Boltzmann constant (m 2 .kg/(s.K)−1), T is the absolute temperature (K), and H is the surface–surface separation distance (m). If Nr is higher than 1, the thermal agitation is negligible compared with the van der Waals attractive forces. For conventional cementitious suspensions without polymers and ultrafine additives, Brownian motion can be neglected compared to the colloidal interactions (Jiao et al., 2021b). The Brownian force causes colloidal particles (also called Brownian particles) to move permanently in a random pattern (Perrin, 1916). This Brownian motion can be slowed down by the high viscosity of the suspending medium through a high diffusive flux. As particle size decreases, Brownian motion gains importance relative to the other forces. A lower limit of Brownian particles is set at 1 nm, below which the size of molecules becomes significant and the homogeneity of the suspending medium can no longer be assumed.

2.1.3 Hydrodynamic force For suspension at flow state, a particle in the suspension experiences a hydrodynamic force tending to facilitate the particle to move. Meanwhile, a viscous drag force is applied to the particle to dissipate the kinetic energy. In the case of a spherical particle in a shear field, the magnitude of the hydrodynamic force is proportional to the square of the particle diameter (d2) (Roussel et al., 2010). The hydrodynamic force FH exerted by the shear field on a particle can be calculated using Eq. (2.3):

FH =

3 πηS d 2γ (2.3) 2

where ηS is the viscosity of the carrier fluid (Pa.s), d is the particle diameter (m), and γ is the shear rate (s−1). 2.2 EFFECT OF COMPOSITIONS ON RHEOLOGY The mixture proportions of cement-based materials vary owing to differences in design requirements. Cement paste usually consists of cementitious materials, water, and chemical admixtures. The rheology of cement paste is influenced by the ratio between the components and the characteristics of each component. In this section, the influences of volume fraction of cementitious materials, characteristics of the interstitial solution, proportion, and properties of cementitious materials, and chemical admixtures on paste rheology are reviewed.

2.2.1 Volume fraction As a suspension, the rheology of cement paste is largely influenced by the volume fraction of the suspending particles. For suspension with a low volume fraction of particles (less than 0.03%), Einstein suggested a simple equation for the viscosity of suspension (Eq. 2.4):

η = ηs (1 + 2.5φ) (2.4)

where ηs is the viscosity of the surrounding fluid and ϕ is the volume fraction of the solid particles. The volume fraction of the cement paste is usually much higher than 0.03%, and

28  Rheology of Fresh Cement-Based Materials

the Krieger–Dougherty model (Eq. 2.5) suggested a better equation for the viscosity of suspension with a high-volume fraction:

 φ  η − ηs  1 + φm  

−[η ]φm

(2.5)

where φm is the maximum volume fraction and [η ] is the intrinsic viscosity. Based on these models, the viscosity of cement paste is influenced by the concentration of cement powders. A higher water-to-cement ratio indicates a lower viscosity. Furthermore, the packing density regarding a high volume concentration and microstructure formation also determine the viscosity of cement paste. Since cement paste is a highly concentrated suspension, the interactions between particles are important because cement particles are not spherical, and the geometry is changed by the progress of hydration. The yield stress of cementitious suspension is also influenced by the volume fraction. According to Zhou et al. (1999), the yield stress of alumina suspensions increased in power law with the growing volume fraction. A yield stress model named Yodel was developed by Flatt and Bowen (2006) to characterize the yield stress of particulate suspension as follows:



τ0 ≅ m

2 A0 a∗ φ (φ − φperc ) (2.6) d 2 H 2 φm (φm − φ )

where m is a pre-factor, which depends on the particle size distribution, d is the particle average diameter, a* is the radius of curvature of the contact points, H is the surface-tosurface separation distance at contact pints, A 0 is the non-retarded Hamaker constant, φ is the solid volume fraction, and φm is the maximum packing fraction of the powder. It shows that apart from the volume fraction of the particles, particle size, particle size distribution, maximum packing, percolation threshold, and interparticle forces together affect the yield stress of the particulate suspensions. The solid volume fraction in cement paste not only influences its rheological parameters but also affects its flow pattern. Roussel et al. (2010) summarized the dominating interactions between particles in cement paste with different volume fractions and shear rate ranges, and the corresponding flow pattern, as shown in Figure 2.1. There exists a percolation volume fraction φperc below which there are no direct contacts nor interactions between the particles and above which the suspension displays yield stress. And there exists a critical volume fraction φdiv above which the yield stress and the viscosity diverge. This critical fraction depends on the degree and/or strength of flocculation of the suspension. Besides, there exists a transition volume fraction of the order of 0.85φdiv which separates suspensions in which the yield stress is mainly dominated by the van der Waals interactions network and concentrated suspensions where the yield stress is mainly due to the direct contacts network. The van der Waals forces dominate the hydrodynamic forces in the low strain rate regime and give rise to a shear-thinning macroscopic behavior. With the increase in strain rate, hydrodynamic forces become more dominant. At high strain rates, however, particle inertia dominates, possibly leading to shear-thickening behavior. On the other hand, unlike many other suspensions, cement paste is thixotropic, and such character is influenced by the volume fraction of the cement paste. Thixotropy is a reversible characteristic of the material being thinned under the condition of disturbance, and the original structure will be restored over time after the disturbance is stopped (Jiao et al., 2021a). It origins from the aggregation between cement particles, which can be influenced

Rheology for cement paste  29

Figure 2.1  Rheo-physical classification of cement suspensions. (Adapted from Roussel et al., 2010.)

by the distance between particles and the volume fraction. According to the experimental work by Lowke (2018), the thixotropic structural build-up of cement paste increases with increasing solid volume fraction. A larger solid volume fraction leads to a shorter distance between particles, and the contributions of colloidal interactions and initial hydration reactions to thixotropy both depend on particle distance.

2.2.2 Interstitial solution Water reducers are indispensable ingredients in flowable cement paste mixtures, enhancing the flowability by dispersing the particles. On the other hand, the non-adsorbed polymers in the interstitial solution may increase the viscosity of the paste, especially in high-performance or ultra-high-performance cement paste, in which the water-to-binder ratio is very low and a massive dosage of superplasticizer is used. As reported by Liu et al. (2017), the relationship between the viscosity of superplasticizer (SP) solution ηs and the concentration of SP in the solution can be fitted in Eq. (2.7). A larger concentration of SP corresponds to a larger viscosity of the solution:

ηs = η0 (1 − φSP ) (2.7) a

Apart from the concentration of the polymers, the ion concentration in the interstitial solution can also influence the rheology of cement paste. On the one hand, the ionic concentration influences the thickness of the electrical double layers of the particles (Hunter, 2001, Cosgrove, 2010). The Debye–Hückel length (1/κ) equation represents the thickness of the double layer (Eq. 2.8) (Overbeek, 1984, Yang et al., 1997).

1 = κ

εε 0 RT (2.8) 2F 2 I

where ε 0 is the permittivity of the vacuum, ε is the dielectric constant (relative permittivity) of the dispersion medium, R is the gas constant, T is the absolute temperature, F

30  Rheology of Fresh Cement-Based Materials

is the Faraday constant, and I is the ionic strength. As the ionic concentration increases, the thickness of electrical double layers decreases, which can result in increased/stronger agglomeration. On the other hand, the sulfate ion concentration in cement paste affects the adsorption behavior of SP on cement particles (Yamada et al., 1998, Yamada et al., 2000). The adsorption of SP decreases in the case of high sulfate ion concentration in the aqueous phase due to the competitive adsorption of sulfate ion and SP on cement particles (Yamada et al., 1998, Yamada et al., 2001).

2.2.3 Cement The mineral composition of ordinary cement clinker is mainly composed of C3S, C2S, C3A, and C4AF. An appropriate amount of gypsum is added to the clinker for grinding cement. The hydration rate and the water demand of reaction of each mineral composition are different. It can be expected that the rheological properties of cement paste may be affected by the mineral compositions of cement. Various ions are released into the water when cement contacts with water, such as SO42−, OH−, Na+, and K+. As mentioned in Section 2.1.5, agglomeration between particles can be enhanced by increasing ion strength, leading to larger yield stress. Besides, a higher hydration rate is expected to a quicker fluidity loss and vice versa. For paste with superplasticizers, these ions may affect the adsorption of superplasticizer onto cement particles and thus affect the rheological properties of cement paste. The fineness of cement (about 350–400 m 2 /kg for ordinary cement) also has a great influence on its rheological properties, because fine cement needs more water for a given flowability, and fine cement hydrates faster than coarser one. The effects of chemical composition and physical characteristics on rheological properties had been studied by many researchers. Hope and Rose (1990) studied the effects of cement composition on water requirement under the constant slump. They found that the required content of mixing water increased for cement with high Al2O3 or C2S contents and decreased for cement with high ignition loss, high carbonate addition, or high C3S content. Havard and Gjorv (1997) examined the influence of the gypsum to hemihydrate ratio of cement on concrete rheology. They stated that for cement with high contents of C3A and alkalis, a reduction in the ratio of gypsum to hemihydrate resulted in a decrease in yield stress but little change in plastic viscosity. For cement with lower contents of C3A and alkalis, the effects of the gypsum to hemihydrate ratio were less pronounced. Furthermore, a reduction in the sulfate content from 3% to 1% caused a decrease in both yield stress and plastic viscosity. Dils et al. (2013) investigated the chemical composition and fineness of cement on rheological properties of ultra-high-performance concrete for a given slump flow. They found that cement with high C3A and specific surface, high alkali content, and a lower content of SO3 gave the worst workability. Chen and Kwan (2012) showed that the influence of superfine cement on rheological properties depends on the water content of cement paste. The addition of superfine cement increased yield stress and apparent viscosity at W/C ≥ 0.24, while at W/C ≤ 0.22, the addition of superfine cement decreased yield stress and apparent viscosity. At lower water content, the addition of fine particles can fill the voids, and it increases the packing density and releases the water between cement particles, and significantly increases the water films coating the particles in the cement paste, consequently improving the rheological properties of cement paste. At higher water content, the influence of increasing superfine cement contents on the water film thickness is not pronounced, and owing to the high specific surface area, superfine cement increased the yield stress and plastic viscosity. From the above analysis, the chemical composition (especially the contents of C3A, C3S, SO3, and alkali), ignition loss,

Rheology for cement paste  31

physical characteristics such as fineness and specific surface area, and the water content have a great influence on rheological properties.

2.2.4 Mineral admixture Mineral admixtures are usually indispensable in modern cement-based materials as supplementary cementitious materials, due to their positive influence on the workability, strength, and durability as well as environmental benefits. This section introduces several most used mineral admixtures including fly ash, ground blast furnace slag, silica fume, and limestone powder, and their effects on the rheology are reviewed. Since the mineral admixtures modify the rheology of cement-based materials by changing the paste phase, and concrete rheology is positively correlated to the rheology of the paste phase, the results of the effects of mineral admixtures on concrete rheology are also included in this section. 2.2.4.1 Fly ash Fly ash, high in SiO2 and Al2O3, and low in CaO, is one type of pozzolanic material, which is sufficiently reactive when mixed with water and Ca(OH)2 at room temperature (Taylor, 1997). Fly ash can delay the early age of hydration, lengthen the setting time, and enhance the long-term strength. Fly ash consists of crystal, vitreous, and a small amount of unburnt carbons. The vitreous includes smooth and spherical-shaped vitreous particles and irregular-shaped and low-porosity small particles. Unburnt carbon presents a loose and porous shape. The chemical and phase compositions depend on the mineral composition of coal, combustion conditions, and collector setup. Physically, fly ash presents as fine particles with particle sizes ranging from 0.4 to 100 μm, low specific gravity (2.0 ~ 2.2 g/cm3), high specific surface area (300 ~ 500 m 2 /kg), and light texture (Chen and Kwan, 2012). The type of fly ash determines the surface texture. For example, the surface of Class C fly ash (FAC) particles appears irregular and cellular (Figure 2.2a), while Class F fly ash (FAF) particles exhibit spherical morphology and smooth surface texture (Figure 2.2b). The addition of fly ash has a great influence on the rheological properties of concrete, as shown in Figure 2.3. Laskar and Talukdar (2008) found that low levels of fly ash could lead to a reduction of yield stress, while a slight increase in yield stress could be observed at the high content of fly ash. Beycioğlu and Aruntaş (2014) indicated that fly ash positively

Figure 2.2 (a and b) SEM images of Class C and F fly ash particles (Ahari et al., 2015b).

32  Rheology of Fresh Cement-Based Materials

Figure 2.3  Several typical results about the effects of fly ash on rheological properties (Grzeszczyk and Lipowski, 1997, Zhang and Han, 2000, Ferraris et al., 2001, Koehler and Fowler, 2004, Park et al., 2005, Laskar and Talukdar, 2008, Vance et al., 2013, Rahman et al., 2014, Ahari et al., 2015b). (Derived from Jiao et al., 2017.)

affected the flowability, passing ability, and viscosity of self-compacting concrete due to the spherical geometry and smooth surface caused a reduction in water demand. Jalal et al. (2013) found that the ball bearing-shaped fly ash particles resulted in an improvement in the rheological properties of fresh self-compacting concrete (SCC), and led to an increase in the slump flow diameter from 800 to 870 mm and a reduction in the T500 values from 1.7 to 1.1 s. However, Park et al. (2005) found that the mixtures without fly ash showed slightly higher yield stress than those with fly ash, and the yield stress and plastic viscosity slightly increased with the increase of fly ash. Rahman et al. (2014) observed that fly ash significantly increased the flocculation rate and plastic viscosity of SCC. The type of fly ash is correlated to the rheological properties of concrete. Ahari et al. (2015a, 2015b) studied the influence of various amounts of FAF and FAC on the rheological properties of SCC. They found that FAF significantly decreased plastic viscosity in comparison to mixtures with FAC. The fineness or particle size distribution of fly ash can also significantly affect workability. Ferraris et al. (2001) and Li and Wu (2005) reached the same conclusion that as the mean size of fly ash particles increased, the slump flow decreased to a certain value and then gradually increased, and the optimum size was about 3 μm. Lee et al. (2003) pointed out that the fluidity of paste increased as the particle size distribution became wider. Fly ash has lower specific gravity than cement. The replacement of cement with the same mass of fly ash could increase the volume of paste, proportionally decrease the cement concentration in the paste, and therefore reduce the number of flocculation cement particle connections, which is called the “dilution effect” (Malhotra and Mehta, 2004, Bentz et al., 2012). The spherical geometry and smooth surface of fly ash particles promote particle sliding and reduce frictional forces among the angular particles, which is called the “ballbearing effect”. This effect can be magnified with the particle size distribution of fly ash slightly coarser than cement due to the increase in separation distance between neighboring particles (Vance et al., 2013). The addition of fine fly ash improves the flowability due to the filling effect. The packing density of paste is improved; the water retained inside the particle flocs is decreased; and consequently, the fluidity of paste is increased. In addition, fly ash can

Rheology for cement paste  33

delay the early age of hydration and lengthen the setting time. It is worth mentioning that the unburnt carbons in fly ash can greatly adsorb superplasticizer molecules or water, even resulting in negative zeta potential (Felekoğlu et al., 2006), which leads to higher viscosity. Stated thus, the replacement level, surface texture, particle size distribution or specific surface area, and content of unburned carbon of fly ash affect the rheological properties of paste by dilution effect, ball-bearing effect, filling effect, and adsorption effect. 2.2.4.2 Ground blast furnace slag Ground blast furnace slag (GBFS), finely granular and almost fully noncrystalline, is a byproduct of the steel-manufacturing industry. The glass content of GBFS is predominated by the rate of quenching. The mass contents of CaO, SiO2 , and Al2O3 are up to 90%. GBFS has hydraulic properties but is much slower than Portland cement. The specific gravity of slag is approximately 2.90, and its bulk density varies in the range between 1200 and 1300 kg/m3. The specific surface area of GBFS measured by the Blaine method is 375–425, 450–550, and 350–450 m 2 /kg in the United Kingdom, the United States, and India, respectively. In China, the specific surface area of GBFS is higher than 450 m 2 /kg (Pal et al., 2003). Since the replacement of cement with slag improves the workability of mixtures and reduces CO2 emissions, GBFS has been widely applied in cement paste, mortar, and concrete. Effects of GBFS on the rheological properties of cement-based materials from literature are shown in Figure 2.4. As can be seen from Figure 2.4, most studies found that the addition of GBFS decreases plastic viscosity, while the effect of GBFS on yield stress is uncertain. Park et al. (2005) indicated that the yield stress decreased and then increased as GBFS increased, but the plastic viscosity decreased with the addition of GBFS. Ahari et al. (2015b) stated that the replacement of Portland cement with GBFS decreased the yield stress and plastic viscosity of the mixtures, regardless of the water-to-binder (w/b) ratio. They also found that an 18% replacement of GBFS in the mixture with a w/b ratio of 0.44 reduced the breakdown area approximately by 10%. In the study by Derabla and Benmalek (2014), the specific surface area of granulated slag and crystallized slags are 228 and 485 m 2 /kg,

Figure 2.4 Several typical results about the effects of slag on rheological properties (Zhang and Zhang, 2002, Koehler and Fowler, 2004, Shi et al., 2004, Park et al., 2005, Boukendakdji et al., 2012, Cao et al., 2012, Derabla and Benmalek, 2014, Ezziane et al., 2014, Tang et al., 2014, Ahari et al., 2015b). (Derived from Jiao et al., 2017.)

34  Rheology of Fresh Cement-Based Materials

respectively, while the activity of these two slags is near to that of cement. Even though the specific surface area of granulated slag is lower than that of cement, the plastic viscosity of the mixture containing granulated slag decreases from 150 to 121 Pa.s due to the absorption of superplasticizer particles. At the same replacement level, the plastic viscosity of the mixture containing crystallized slag decreases from 150 to 91 Pa.s. However, the addition of GBFS can increase the plastic viscosity in some cases. Tattersall (1991) reported that for the mixtures with low cement content (200 kg/m3), the addition of slag reduced yield stress and increased plastic viscosity, which was used at replacement levels of 40% and 70%. Tang et al. (2014) studied the rheological properties of cement paste with a high volume of ground slag under different slump flow conditions by varying the dosage of superplasticizer. They found that the pastes with the ground slag had higher plastic viscosity, poorer stability, and lower velocity of flow, compared to ordinary cement paste under the same slump flow condition. These results can be attributed to the fact that a high volume of ground slag with a high specific surface area requires a larger amount of water than cement. Generally, GBFS has a high specific surface area and a high chemical activity, which may have positive or negative effects on the rheological properties of cement-based materials. The rheological properties of cement concrete are improved for the following reasons: entrapped water within cement particles may be released by the micro-filling effect of fine GBFS particles, and a high specific area may lead to more adsorption of superplasticizer. However, high specific surface area and high chemical activity of ground blast furnace slag require a large amount of water than cement particles, and thus the rheological properties may be decreased. As a result, the influence of GBFS on rheological properties depends on the replacement level, chemical composition, specific surface area or fineness, adsorption effect, and water requirement. 2.2.4.3 Silica fume Silica fume is the by-product of electric arc furnace-produced silica metal and silica alloys. Silica fume is an extremely fine powder. The average particle size is about 0.1–0.3 μm, and the particles below 0.1 μm are more than 80%. The specific surface area of silica fume is 20,000–28,000 m 2 /kg, which is 80–100 times and 50–70 times higher than that of cement and fly ash, respectively. The particles are round and tend to be agglomerated. Therefore, superfine silica fume particles can fill the voids between other particles, which improves the gradation, increases the packing density of cementitious materials, and even has a lubrication effect. The high specific surface area of silica fume could even adsorb superplasticizer molecules with multi-layers (Nehdi et al., 1998, Park et al., 2005, Laskar and Talukdar, 2008). Silica fume, high in SiO2 (85%–96%) with very fine vitreous particles, has higher chemical activity than fly ash. The high fineness and high chemical activity of silica fume can increase the water demand and interparticle friction (Collins and Sanjayan, 1999, Nanthagopalan et al., 2008, Benaicha et al., 2015). Effects of silica fume on the rheological properties of cement-based materials from literature are shown in Figure 2.5. As can be seen from Figure 2.5, many studies found that the addition of silica fume increased both yield stress and plastic viscosity, and reduced the fluidity of cement-based materials. The addition of silica fume also significantly increased the flocculation rate (Rahman et al., 2014, Ahari et al., 2015b). As a result, silica fume can be used in the concrete which requires high uniformity and cohesiveness, e.g. pumping concrete, underwater concrete, and shotcrete, as an inorganic viscosity-modifying agent. However, the addition of silica fume may have different effects on the rheological properties. Zhang and Han (2000) stated that silica fume could decrease the viscosity and yield stress of cement paste. Ahari et al. (2015a, 2015b) found that the incorporation of silica

Rheology for cement paste  35

Figure 2.5  Several typical results about the effects of silica fume on rheological properties (Zhang and Han, 2000, Ferraris et al., 2001, Koehler and Fowler, 2004, Park et al., 2005, Laskar and Talukdar, 2008, Rahman et al., 2014, Ahari et al., 2015b, Lu et al., 2015). (Derived from Jiao et al., 2017.)

Figure 2.6  Effect of replacement of silica fume on rheological parameters (Nanthagopalan et al., 2008).

fume reduced the plastic viscosity and increased the yield stress and breakdown area. Yun et al. (2015) found that silica fume led to a remarkable increase in flow resistance, while it slightly reduced plastic viscosity. The addition of silica fume could obtain a different result with different superplasticizer types and water-to-binder ratios. Laskar and Talukdar (2008) found that silica fume increased the initial yield stress of fresh concrete in the case of polycarboxylate-based superplasticizer and decreased it in the case of sulfonated naphthalene polymer. Effects of water-to-powder ratio (w/p) and silica fume on yield stress and plastic viscosity are shown in Figure 2.6. As can be seen from Figure 2.6, both yield stress and plastic viscosity show a higher increment with the decrease of water-to-powder ratio.

36  Rheology of Fresh Cement-Based Materials

Therefore, it is necessary to understand the interaction between silica fume and superplasticizer type or content when investigating the effect of silica fume on the rheology of ordinary concrete. In summary, specific surface area and chemical activity are the most important factors affecting the rheological properties because of micro-filling effect, lubricating effect or friction effect, and adsorption effect. Furthermore, the efficiency of silica fume can be influenced by the superplasticizer type or content and water-to-binder ratio. 2.2.4.4 Limestone powder The inert limestone powder is a kind of high-quality and cheap mineral admixture. Its main component is calcium carbonate, CaCO3. The surface of limestone particles is irregular and rough (Ma et al., 2013b), which increases the adhesion and friction between cement particles. The mean particle size of limestone powder used in concrete is from below 1 μm to more than several tens of μm. Limestone particles have a high adsorption capacity of superplasticizer, which increases the dispersing ability of the concrete system. Besides, limestone particles are often ground in finer powders than cement. Consequently, more water is needed for paste with the fine limestone powder. The influences of limestone powder on rheological properties from literature are summarized in Figure 2.7. Some researchers found that the incorporation of limestone powder increased the yield stress and plastic viscosity, leading to a reduction in the workability of concrete, while others showed that the addition of limestone powder resulted in a decrease in yield stress and plastic viscosity. In addition, Rahman et al. (2014) stated that increasing amounts of limestone powder led to an increase in the flocculation rate significantly. For instance, the structuration rate of concrete with 15% limestone powder is about 1.5 times higher than that of reference concrete. The influence of limestone powder on rheological properties depends on the particle packing and water demand, which could be transferred into specific surface area or particle size distribution (Vance et al., 2013). Ma et al. (2013a) noted that yield stress and plastic viscosity increased with decrease in the particle size of limestone powder. Vance et al. (2013) found that as the particle size of limestone powder increased from 0.7 to 15 μm, there was a decrease in both yield stress and plastic viscosity. They stated that particles finer than the ordinary Portland cement usually increase the yield

Figure 2.7 Several typical results about the effects of limestone powder on rheological properties (Yahia, 1999, Zhang and Han, 2000, Vance et al., 2013, Derabla and Benmalek, 2014, Ezziane et al., 2014, Rahman et al., 2014). (Derived from Jiao et al., 2017.)

Rheology for cement paste  37

stress and plastic viscosity, while the addition of particles coarser than cement induces an opposite effect. This is because the finer particles reduce the particle spacing and increase interparticle contacts, whereas the coarser particles could increase particle spacing and reduce the shear resistance of the suspension. Furthermore, larger limestone particles exhibited a decrease in the area of water films and provided more extra water (Uysal and Yilmaz, 2011), and thus reduced the possibility of flocculation. Although the effect of particle size distribution is very important, the production methods of limestone powder cannot be ignored. Felekoğlu et al. (2006) considered that the finer limestone filler did not significantly change the viscosity, while the coarse limestone filler was effective in increasing the viscosity. The finer limestone filler is a filtration system by-product of crushed stone production. However, the coarse limestone powder is a special production of whitened limestone filler with lower adsorptivity. In a word, the effect of limestone powder on rheological properties can be ascribed to the morphologic effect, filling effect, and adsorption effect by particle size distribution and production methods of limestone powder. 2.2.4.5 Ternary binder system Ternary binder system is a good way to enhance various properties of cement-based materials, including workability, rheological properties, strength, durability, cost, and CO2 emissions. This section discusses the rheological properties of ternary blends containing fly ash and slag or silica fume or limestone. Laskar and Talukdar (2008) observed that the rheological parameters of concrete with a ternary binder system lay in between the values with the single mineral admixtures at each replacement level. Tattersall and Banfill (1983), Park et al. (2005), and Gesoğlu and Özbay (2007) observed the same phenomena. Kashani et al. (2014) investigated the rheological behavior of cement-blast furnace slag-fly ash ternary pastes. They showed that the width of the particle size distribution was the key parameter controlling the yield stress of ternary pastes. As a result, even a low volume of fly ash also has a significant effect on workability, and any other mineral additives with a broad particle size distribution may have a comparable effect. Vance et al. (2013) found that in ternary pastes containing limestone with low fly ash contents (5%), the plastic viscosity increased with the replacement of fine limestone and remained unchanged for coarse limestone. However, at higher fly ash contents (10%), the yield stress decreased even for pastes containing fine limestone. Although the particle spacing reduces and the number of interparticle contacts increases with the addition of fine limestone, the ball-shaped fly ash separates the solid grains, compensating for the positive effect of fine particle additions, thus reducing the yield stress and plastic viscosity. The addition of ultrafine mineral admixtures such as silica fume can increase the packing density, fill the voids between cement particles, and release the water entrapped in agglomerate structure to form excess water films for lubrication. But at the same time, the water film thickness will be thinned down due to its large surface area. By adding a cementitious material whose fineness lies in between cement and silica fume, such as fly ash, the water entrapped in the agglomerate structure can be released without excessively increasing the surface area (Li and Kwan, 2014). In this way, a larger water film thickness and better flowability can be obtained. Moreover, the addition of mineral admixtures with a broader particle size distribution can also increase the particle packing and the excess water, and therefore can efficiently separate the cement particles. In this case, the intensity of colloidal interaction between particles will drop substantially, leading to the destruction of the percolated network of particles, which will then result in yield stress reduction (Kashani et al., 2014).

38  Rheology of Fresh Cement-Based Materials

2.2.5 Chemical admixtures With the wide application of high-performance concrete, admixtures have become an indispensable part of concrete. The types of chemical admixtures have a great influence on the rheological properties of fresh concrete. 2.2.5.1 Superplasticizer Due to economic and technical benefits, superplasticizers are almost used for all modern concretes. The yield stress and viscosity of cement-based materials can be dramatically decreased by the addition of superplasticizer under the action of electrostatic or/and steric hindrance effects. The efficiency of superplasticizer depends on the adsorption of superplasticizer molecules onto cement particles and the repulsive force developed by the adsorbed molecules. For the new generation of superplasticizer, it seems that the steric hindrance effect becomes dominant over the electrostatic effect (Flatt and Bowen, 2006). With the increase of superplasticizer, the average surface-to-surface separation distance increases, the colloidal interaction between particles decreases, and therefore the yield stress of cementitious materials decreases (Perrot et al., 2012). Beyond that, the addition of superplasticizer leads to a higher surface coverage by polymers—causing an increase in effective layer thickness and a reduction in the maximum attraction between the particles, thus decreasing the number of sites available for nucleation, and increasing the bridging distance between the particles—and consequently improves the rheological properties of fresh concrete (Lowke et al., 2010, Kwan and Fung, 2013). The efficiency of superplasticizer depends on its type and structure. Figure 2.8 demonstrates the rheological properties of SCC made with a different type of superplasticizer. It can be observed from Figure 2.8 that the concrete mixture with polycarboxylate-based superplasticizer has a higher yield stress and a lower plastic viscosity compared to those with naphthalene sulphonate-based superplasticizer at the same slag content. Papo and Piani (2004) found that modified polyacrylic was the most effective superplasticizer to improve the rheological properties of cement paste among three types of superplasticizers based on melamine resin, modified lignosulphonate, and modified polyacrylate. The naphthalene

Figure 2.8 Effect of superplasticizer type on rheological properties of SCC (Boukendakdji et al., 2012): (a)  yield stress and (b) plastic viscosity. SP1: a polycarboxylate-based superplasticizer; SP2: a naphthalene sulphonate-based superplasticizer.

Rheology for cement paste  39

superplasticizer is beneficial for dispersing the cement, with a low air-entraining effect and slump retention effect. The polycarboxylate superplasticizer consisting of a backbone of polyethylene, grafted chains of PEO, and carboxylic groups as adsorbing functional groups, has the potential to disperse cement particles, restrain the slump loss, and shorten the setting time (Zuo et al., 2004, Hanehara and Yamada, 2008, Boukendakdji et al., 2012). Toledano-Prados et al. (2013) found that liquid polycarboxylate proved more effective than solid polycarboxylate. Different admixtures that had the same main chain and same polymer structure but different molecular weight and different side chain density of carboxylic acid groups have a great effect on the rheological properties of SCC (Mardani-Aghabaglou et al., 2013). The yield stress values of SCC mixtures were only affected by the dosage of superplasticizer, while increasing both the molecular weight and side chain density of carboxylic acid groups of admixture increased the plastic viscosity values due to the increase in steric hindrance. 2.2.5.2 Viscosity-modifying agent Viscosity-modifying agents (VMA) are relatively new admixtures used to improve the rheology of concrete with higher uniformity and better cohesiveness. Incorporation of VMA in concrete can reduce the risk of separation of heterogeneous concrete and obtain stable concrete for underwater repair, curtain walls, and deep foundation walls. The mode of action of VMA depends on the type and concentration of the polymer in use. The mechanism of action with welan gum and cellulose derivatives can be classified into three categories, i.e. adsorption effect, association effect, and intertwining effect (Khayat, 1995). The adsorbed cellulose ether can slow down the nucleation of calcium silicates, generate repulsive steric forces to replace van der Waals attractive forces, and then form a new interaction network that could bridge the cement grains (Brumaud et al., 2014). The inorganic VMA with a high surface area increases the content of fine particles and the water-retaining capacity of paste, thereby the thixotropy. The concrete modified with a VMA exhibits high plastic viscosity, high yield value, and shear-thinning behavior. The efficiency of VMA depends on its type and concentration (Khayat, 1995). Schmidt et al. (2010) stated that the influence of VMA on yield stress was much stronger than the effect of polycarboxylate ether superplasticizer (PCE) at 20°C, and the mixtures with VMA based on modified potato starch show better performance than those with VMA based on diutan gum, irrespective of the PCE type. Yun et al. (2015) found that VMA based on hydroxypropyl methyl cellulose tended to reduce the flow resistance and increase the plastic viscosity of high-performance wet-mix shotcrete, worsening the shootability and pumpability. Leemann and Winnefeld (2007) showed that VMA based on polysaccharide caused the highest increase in yield stress, while VMA based on microsilica had the lowest at an equivalent plastic viscosity. Assaad et al. (2003) showed that the addition of VMA significantly increased the thixotropy for high-flowability concrete. Brumaud et al. (2014) found that the critical deformation increased with the VMA dosages. One possible explanation is that the number of bridging grains increases and therefore the probability of large interparticle relative displacement increases. Another possible explanation comes from the fact that the hydrophobic interactions of high VMA concentrations could form aggregates of large size, which could tolerate higher stretching (Bülichen and Plank, 2012, Bülichen et al., 2012). Different types of VMA may have the same effect. Benaicha et al. (2015) showed that the concrete made with 10% of silica fume and the one with 0.1% VMA had the same characteristics in resistance of sieve segregation, plastic viscosity, yield stress, and even mechanical properties. They believed that VMA could be replaced by silica fume, depending on the availability of materials.

40  Rheology of Fresh Cement-Based Materials

VMA is often used in combination with superplasticizer to improve rheological properties and mechanical strength. The presence of two dissimilar chemicals can lead to a number of issues related to incompatibility, which affects the properties of concrete. Khayat (1995) stated that cellulose derivatives were always used in conjunction with melamine-based superplasticizer because of their incompatibilities with a naphthalene-based superplasticizer. Prakash and Santhanam (2006) indicated that the flow properties of pastes produced with combinations of superplasticizer based on sulfonated naphthalene formaldehyde or polycarboxylic ether and VMA based on welan gum were satisfactory. The only difference was that the polycarboxylic ether–welan gum combination showed the evidence of thixotropy, but the same was not observed for sulfonated naphthalene formaldehyde–welan gum combinations. In addition, the rheological behavior of mixtures with VMA is in general dependent upon shear stress. Bouras et al. (2012) found that at sufficiently low shear stress, the mixtures with VMA exhibited shear thinning, while at relatively high shear stress, the pastes became shear thickening. At low shear stress, the flow of paste induces the deflocculation of solid particles and VMA polymer disentanglement and alignment. The analysis above shows that the efficiency of the viscosity-modifying agent depends on the type and concentration of VMA, the type of superplasticizer, and the applied shear stress. 2.2.5.3 Air-entraining agent The air-entraining agent has been widely used to improve resistance to freezing and thawing damage, and to a lesser extent, the workability of concrete. The air-entraining agent is a mixture of various surfactants (Du and Folliard, 2005). The chemical nature of the air-entraining agent contains a hydrophilic head, having a strong attraction for water, and hydrophobic tails, having repulsion for water. From the rheological point of view, entrained air bubbles play a role in lubrication, and increasing paste volume, together with the character of the air-entraining agent, affects the consistency of cement-based materials. There are many studies about the effect of the air-entraining agent on rheological properties. He et al. (2011) found that consistency of mortar showed a slightly increasing trend with the increase of air-entraining amount, although the value of consistency, in general, was not high. Yun et al. (2015) found that the use of an air-entraining agent tended to reduce both flow resistance and torque viscosity, effectively improving the pumpability of high-performance wet-mix shotcrete, while the reduction rate of torque viscosity decreased with the increased air content. Tattersall and Banfill (1983) demonstrated that flow resistance and torque viscosity continued to decrease until the air content reached 5% and then stabilized beyond that point. This result can be contributed to the lubrication effect and volume effect of air bubbles. However, Carlsward et al. (2003) and Wallevik and Wallevik (2011) revealed that the addition of an air-entraining agent strongly reduced the plastic viscosity, while the effect on yield stress was not significant. Struble and Jiang (2005) indicated that the yield stress increased and the plastic viscosity decreased with increasing air content, which was also reported by Rahman and Nehdi (2003). An explanation for the increase in yield stress is proposed that air bubbles are attracted to cement particles to form bubble bridges and thus increase bonding between particles. Once the bubble bridges are broken and paste can flow, the air bubbles reduce plastic viscosity, apparently acting as a lubricant. Thus, there is competition between these two actions: without shear, the bubbles act as flocculating particles to increase yield stress, while with shear, the bubbles act as lubricant to reduce plastic viscosity (Edmeades and Hewlett, 1998). Moreover, the rheological behavior of suspensions with bubbles also depends on the shear rate due to the fact that the bubbles are generally stiff compared to the suspending fluid and thus do not store any energy (Ducloué et al., 2015).

Rheology for cement paste  41

2.3 EFFECT OF TEMPERATURE ON RHEOLOGY Temperature usually affects the rheological properties of cement-based materials by changing the actions of admixtures and the interaction between cementitious material particles. Moreover, temperature impacts the time-dependent rheological properties by changing the flocculation rate of cementitious material particles and the kinetics of cement hydration. For the cement paste containing water-reducing agent, according to Fernàndez-Altable and Casanova (2006), the viscosity decreased with the increasing temperature, while the static and dynamic yield stress increased with the elevated temperature. However, the influence of temperature on yield stress was only obvious for paste with a low dosage of water reducer. For cement pastes with a high dosage of water reducer, the dependence of yield stress on temperature was found. Nawa et al. (2000) found that the fluidity of cement paste did not change monotonically with increasing temperature. They found that the fluidity was smallest at 20°C, while the fluidity loss was larger at a higher temperature. On the one hand, the increase in temperature will increase the adsorption of polycarboxylate superplasticizer, enhance the steric resistance between cementitious material particles, and improve the fluidity of paste. On the other hand, the increase in temperature promotes the hydration process, and some polymers will be buried in the laminates of hydration products, thus reducing their performance. On the performance of superplasticizer at different temperatures, Kong et al. (2013) studied the rheological properties of Portland cement pastes made with polycarboxylate superplasticizers. According to their work, a higher temperature can lead to a more significant amount of adsorbed polycarboxylate ester/ether on the cement surface and a lower amount of free water in fresh cement pastes, because of the higher hydration rate of cement. Yamada et al. (1999) attributed the low dispersibility of polycarboxylate superplasticizer at a lower temperature to the high sulfate ion concentration in mixing water. Compared with the effect of temperature on the initial rheological properties, the effect of temperature on the time-varying rheological properties is more significant. Al Martini and Nehdi (2009), Al Martini (2008), and Nehdi and Al Martini (2009) studied the effect of high temperature on the rheology of cement paste and its change with time. It was found that such influences were affected by the superplasticizer (SP) dosage. When the SP dosage was below the saturation level, the increase in yield stress over time at high temperature was obviously larger, while at the dosage higher than the saturation level, the influence of temperature on yield stress evolution over time was less obvious. A similar principle was also applicable to the effect of temperature and time on plastic viscosity. In the studies of Petit et al. (2006, 2009, 2007), coupled effects of temperature and time on rheological parameters were studied, and correlations between hydration time and rheological parameters were established. They found that both yield stress and plastic viscosity vary linearly with material temperature and elapsed time for mixtures made with polymelamine or polynaphtalene superplasticizer. However, for flowable mortar with polycarboxylic superplasticizer, rheological properties can be influenced by both mixture proportioning and temperature. Besides, the effect of temperature on the variation of viscosity is more obvious in the mortar with a w/b of 0.42 than that of 0.53. 2.4 EFFECT OF SHEARING ON RHEOLOGY As is well known, the operation processes of modern engineering applications of concrete include mixing, transporting, pumping, and formwork casting. At each process, fresh concrete is subjected to different shear rates. For example, the process of mixing can provide a

42  Rheology of Fresh Cement-Based Materials

“most deflocculated state” for the fresh concrete due to its high shear rates (Roussel, 2006). The applied shear rate during transport is associated with the rotational speed, drum volume, and rheological properties of the concrete itself, and its value varies from 1 to 8 s−1 (Wallevik and Wallevik, 2017). For the process of pumping, the fresh concrete is subjected to very high shear stress and high pressure at the same time. As a result, the yield stress and air content are markedly increased and the viscosity of fresh concrete is decreased (Secrieru et al., 2018, Shen et al., 2021). However, the only source of applied shear stress of concrete after casting into formwork is gravimetric force, and thus the applied shear rate is usually less than 0.1 s−1 (Papanastasiou, 1987). It is worth mentioning that the shear rate experienced by the cement pastes inside fresh concrete depends on the nature of the concrete (Roussel, 2006). The operation process, i.e. shear history, plays an important role in the evolutions of rheological properties and stability of fresh concrete over time. The dispersion and agglomeration of particles in suspensions are significantly dependent on the applied shear rates. As mentioned before, the operations of mixing and pumping with high shear rates can almost totally break down the network structures of cement paste. Although paste in concrete suffers higher shear rates than pure paste under the same shear conditions, Helmuth et al. (1995) and Ferraris et al. (2001) pointed out that the rheological properties of cement paste prepared with a high-speed mixer were comparable with that of paste mixed in concrete. Williams et al. (1999) also mentioned that the well-mixed cement pastes contained few agglomerate structures and possessed low plastic viscosity. However, Han and Ferron (2015, 2016) found that the rheological properties of cement paste worsened once the mixing intensity was higher than a threshold value due to the agglomerates with a larger mean chord size. Recently, Mostafa et al. (2015) evaluated the effectiveness of additional shearing on the dispersion state of cement suspensions. They found that applying a rotational shearing with the shear rate corresponding to the transition shear rate could significantly improve the dispersing degree of fresh cement suspensions. The rheological properties of concrete are relevant to the dispersion state of the paste matrix (Tregger et al., 2010), and it is influenced by the shear history (Ferron et al., 2013). Generally, a higher shear rate can break down the flocs of binders, thus improving the flowability of the paste (Jiao et al., 2018). However, there are opposite findings as well. As reported by Helmuth (1980), a very high shear rate can degrade the fluidity of cement paste. Han and Ferron (2016) found that weaker but larger cement agglomerates can form in the paste under high mixing rates, and thus the rheological properties of cement paste can increase after exceeding a threshold mixing speed. On the other hand, shear history can also influence the aggregation and breakage kinetics of the paste (Ferron et al., 2013). Ferron et al. (2013) found that the time scale needed for aggregation is longer than the breakdown of the flocculation structure. Ma et al. (2018) found that longer shearing duration can induce a more dispersed structure and decrease the static yield stress and the rebuild rate of static yield stress. Besides, increasing the pre-shearing time can increase the storage modulus value, leading to a more percolated C–S–H network (Ma et al., 2018). It is important to understand how the action of shearing influences the rheological behavior of cement-based materials to evaluate their rheology in the real situation more precisely. 2.5 EFFECT OF PRESSURE ON RHEOLOGY Cement-based materials are pressured when they flow through pipes driven by pumps or in the formworks, and their rheological properties can be different under pressure. The influence of pressure on the rheological properties of cement pastes, which assumably represented the lubricating layer that forms along with the profile of concrete during pumping,

Rheology for cement paste  43

was evaluated using a rotational rheometer with a high-pressure cell in the work of Kim et al. (2017). Cement pastes with water-to-cement ratios ranging from 0.35 to 0.6 were tested according to a protocol designed to simulate the conditions of an actual pumping process based on field tests. The shear rates, shearing durations, and pressure levels from 0 to 30 MPa were experimentally simulated. The test results indicated that below a certain water-to-cement ratio (0.40), elevated pressures lead to changes in the rheological properties, while changes were negligible when the ratio was above this threshold. Further, at low water-to-cement ratios, the thixotropy of cement pastes can reverse into rheopexy after pressurization.

2.6 SUMMARY Cement paste can be regarded as a suspension system containing particles from microscopic to macroscopic dimensions. Interactions between particles, including colloidal interaction, Brownian forces, and hydrodynamic forces, are the origin governing the rheology of cement paste. Cement paste is thixotropic, and its rheology is time-dependent. The rheological properties of cement paste are not only influenced by mixture proportion but also by processing procedures. Mixture proportion factors, i.e. volume fraction of cementitious materials, characteristics of the interstitial solution, proportion and properties of cementitious materials, and chemical admixtures, can change the rheological behavior of cement paste by modifying the interactions between particles. The mineral composition of cement influences the hydration rate and water demand in paste, and thus affects the paste rheology. The chemical environment in the interstitial solution can also be altered, and it can further change the agglomeration of cement particles. Mineral admixtures, such as fly ash, slag powder, silica fume, and limestone powder, mainly influence the rheological properties by changing the physical properties of particles including specific surface area, particle size distribution, and particle shape. Chemical admixtures including superplasticizer, VMA, and the air-­ entraining agent can change the rheology by directly influencing particle interactions. Processing factors, i.e. temperature, shearing, and pressure, alter the rheology of cement paste by changing the flocculation and chemical hydration in cement paste. Increasing temperature can increase superplasticizer adsorption and promote cement hydration in cement paste. Shearing effect can break part of the flocs and change the dispersion state. Furthermore, additional dissolution and superplasticizer adsorption may occur on newly generated surfaces of cement particles, promoting cement hydration and SP adsorption. The knowledge of the rheology of cement paste under pressure remains pretty limited.

REFERENCES Ahari, R. S., Erdem, T. K., and Ramyar, K. (2015a). “Effect of various supplementary cementitious materials on rheological properties of self-consolidating concrete.” Construction and Building Materials, 75, 89–98. Ahari, R. S., Erdem, T. K., and Ramyar, K. (2015b). “Thixotropy and structural breakdown properties of self consolidating concrete containing various supplementary cementitious materials.” Cement and Concrete Composites, 59. doi:10.1016/j.cemconcomp.2015.03.009. Al-Martini, S. (2008). “Investigation on rheology of cement paste and concrete at high temperature,” Doctoral dissertation, University of Western Ontario. Available at: http://search.proquest.com. ezaccess.library.uitm.edu.my/docview/305111374?accountid=42518.

44  Rheology of Fresh Cement-Based Materials Al Martini, S., and Nehdi, M. (2009). “Coupled effects of time and high temperature on rheological properties of cement pastes incorporating various superplasticizers.” Journal of Materials in Civil Engineering, 21(8), 392–401. 10.1061/(ASCE)0899-1561(2009)21:8(392). Assaad, J., Khayat, K. H. and Mesbah, H. (2003). “Assessment of thixotropy of flowable and selfconsolidating concrete.” ACI Materials Journal, 100(2), 99–107. doi:10.14359/12548. Banfill, P. F. G. (1979). “A discussion of the papers “rheological properties of cement mixes” by M. Daimon and DM Roy.” Cement and Concrete Research, 9(6), 795–796. Benaicha, M., et al. (2015). “Influence of silica fume and viscosity modifying agent on the mechanical and rheological behavior of self compacting concrete.” Construction and Building Materials, 84, 103–110. Bentz, D. P., et al. (2012). “Influence of particle size distributions on yield stress and viscosity of cement-fly ash pastes.” Cement and Concrete Research, 42(2), 404–409. doi:10.1016/j. cemconres.2011.11.006. Beycioğlu, A., and Aruntaş, H. Y. (2014). “Workability and mechanical properties of self-­compacting concretes containing LLFA, GBFS and MC.” Construction and Building Materials, 73, 626–635. Boukendakdji, O., Kadri, E.H., and Kenai, S. (2012). “Effects of granulated blast furnace slag and superplasticizer type on the fresh properties and compressive strength of self-compacting concrete.” Cement and Concrete Composites, 34(4), 583–590. Bouras, R., Kaci, A., and Chaouche, M. (2012). “Influence of viscosity modifying admixtures on the rheological behavior of cement and mortar pastes.” Korea-Australia Rheology Journal, 24(1), 35–44. Brumaud, C., et al. (2014). “Cellulose ethers and yield stress of cement pastes.” Cement and Concrete Research, 55, 14–21. Bülichen, D., Kainz, J., and Plank, J. (2012). “Working mechanism of methyl hydroxyethyl cellulose (MHEC) as water retention agent.” Cement and Concrete Research, 42(7), 953–959. Bülichen, D., and Plank, J. (2012). “Mechanistic study on carboxymethyl hydroxyethyl cellulose as fluid loss control additive in oil well cement.” Journal of Applied Polymer Science, 124(3), 2340–2347. Cao, M., Zhang, C., and Han, L. (2012). “Experimental study on the rheological properties of fresh mineral powder concrete based on two-point method.” Hunningtu (Concrete), (01), 138–141. Carlsward, J., et al. (2003). “Effect of constituents on the workability and rheology of self-­compacting concrete.” Proceeding of the Third International RILEM Conference on SCC, Reykjavik, Island, Proceedings PRO, pp. 143–153. Chen, J. J., and Kwan, A. K. H. (2012). “Superfine cement for improving packing density, rheology and strength of cement paste.” Cement and Concrete Composites, 34(1), 1–10. Collins, F., and Sanjayan, J. G. (1999). “Effects of ultra-fine materials on workability and strength of concrete containing alkali-activated slag as the binder.” Cement and Concrete Research, 29(3), 459–462. Cosgrove, T. (2010). Colloid science: Principles, methods and applications. John Wiley & Sons, New Jersey, United States. Derabla, R., and Benmalek, M. L. (2014). “Characterization of heat-treated self-compacting concrete containing mineral admixtures at early age and in the long term.” Construction and Building Materials, 66, 787–794. Dils, J., Boel, V., and De Schutter, G. (2013). “Influence of cement type and mixing pressure on air content, rheology and mechanical properties of UHPC.” Construction and Building Materials, 41, 455–463. Du, L., and Folliard, K. J. (2005). “Mechanisms of air entrainment in concrete.” Cement and Concrete Research, 35(8), 1463–1471. Ducloué, L., et al. (2015) “Rheological behaviour of suspensions of bubbles in yield stress fluids.” Journal of Non-Newtonian Fluid Mechanics, 215, 31–39. Edmeades, R. M., and Hewlett, P. C. (1998) “Cement admixtures”, in Lea’s chemistry of cement and concrete, edited by Hewlett, P. and Liska, M. Butterworth-Heinemann, Oxford, United Kingdom, pp. 841–905.

Rheology for cement paste  45 Ezziane, K., Ngo, T.-T., and Kaci, A. (2014). “Evaluation of rheological parameters of mortar containing various amounts of mineral addition with polycarboxylate superplasticizer.” Construction and Building Materials, 70, 549–559. Felekoğlu, B., et al. (2006). “The effect of fly ash and limestone fillers on the viscosity and compressive strength of self-compacting repair mortars.” Cement and Concrete Research, 36(9), pp. 1719–1726. Fernàndez-Altable, V., and Casanova, I. (2006). “Influence of mixing sequence and superplasticiser dosage on the rheological response of cement pastes at different temperatures.” Cement and Concrete Research, 36(7), 1222–1230. doi:10.1016/j.cemconres.2006.02.016. Ferraris, C. F., Obla, K. H., and Hill, R. (2001). “The influence of mineral admixtures on the rheology of cement paste and concrete.” Cement and Concrete Research, 31(2), 245–255. Ferron, R. D., et al. (2013). “Aggregation and breakage kinetics of fresh cement paste.” Cement and Concrete Research, 50, 1–10. doi:10.1016/j.cemconres.2013.03.002. Flatt, R. J. (2004a). “Dispersion forces in cement suspensions.” Cement and Concrete Research, 34(3), 399–408. doi:10.1016/j.cemconres.2003.08.019. Flatt, R. J. (2004b). “Towards a prediction of superplasticized concrete rheology.” Materials and Structures, 37(5), 289–300. Flatt, R. J., et al. (2009). “Conformation of adsorbed comb copolymer dispersants.” Langmuir, 25(2), 845–855. Flatt, R. J., and Bowen, P. (2003). “Electrostatic repulsion between particles in cement suspensions: Domain of validity of linearized Poisson–Boltzmann equation for nonideal electrolytes.” Cement and Concrete Research, 33(6), 781–791. Flatt, R. J., and Bowen, P. (2006). “Yodel: A yield stress model for suspensions.” Journal of the American Ceramic Society, 89(4), 1244–1256. doi:10.1111/j.1551–2916.2005.00888.x. Gesoğlu, M., and Özbay, E. (2007). “Effects of mineral admixtures on fresh and hardened properties of self-compacting concretes: Binary, ternary and quaternary systems.” Materials and Structures, 40(9), 923–937. Grzeszczyk, S., and Lipowski, G. (1997). “Effect of content and particle size distribution of highcalcium fly ash on the rheological properties of cement pastes.” Cement and Concrete Research, 27(6), 907–916. Han, D., and Ferron, R. D. (2015). “Effect of mixing method on microstructure and rheology of cement paste.” Construction and Building Materials, 93. doi:10.1016/j.conbuildmat.2015.05.124. Han, D., and Ferron, R. D. (2016). “Influence of high mixing intensity on rheology, hydration, and microstructure of fresh state cement paste.” Cement and Concrete Research, 84, 95–106. doi:10.1016/j.cemconres.2016.03.004. Hanehara, S., and Yamada, K. (2008). “Rheology and early age properties of cement systems.” Cement and Concrete Research, 38(2), 175–195. Havard, J., and Gjorv, O. E. (1997). “Effect of gypsum-hemihydrate ratio in cement on rheological properties of fresh concrete.” Materials Journal, 94(2), 142–146. He, Z. M., Liu, J. Z., and Wang, T. H. (2011). “Influence of air entraining agent on performance of inorganic thermal insulating mortar”, in Applied Mechanics and Materials, edited by Sun, D.Y., Sung, W.P., and Chen, R. Trans Tech Publ, Stafa-Zurich, Switzerland, pp. 490–493. Helmuth, R. A. (1980). “Structure and rheology of fresh cement paste.” 7th International Congress on the Chemistry of Cement. Helmuth, R. A., et al. (1995). “Abnormal concrete performance in the presence of admixtures”, Portland Cement Association. Available at: https://trid.trb.org/view/460037 (Accessed: 22 August 2018). Hiemenz, P. C., and Rajagopalan, R. (1997). Principles of colloid and surface chemistry, 3rd ed. Marcel Dekker, New York. Hope, B. B., and Rose, K. (1990). “Statistical analysis of the influence of different cements on the water demand for constant slump. Properties of fresh concrete.” Proceedings of the Coll, RILEM, Chapman and Hall, p. 179e186. Hunter, R. J. (2001). Foundations of colloid science. Oxford University Press, Oxford, United Kingdom.

46  Rheology of Fresh Cement-Based Materials Jalal, M., Fathi, M., and Farzad, M. (2013). “Effects of fly ash and TiO2 nanoparticles on rheological, mechanical, microstructural and thermal properties of high strength self compacting concrete”, Mechanics of Materials, 61, pp. 11–27. Jiao, D., et al. (2017). “Effect of constituents on rheological properties of fresh concrete-A review.” Cement and Concrete Composites, 83, pp. 146–159. doi:10.1016/j.cemconcomp.2017.07.016. Jiao, D., De Schryver, R., et al. (2021a). “Thixotropic structural build-up of cement-based materials: A state-of-the-art review.” Cement and Concrete Composites, 122(February), 104152. doi:10.1016/j.cemconcomp.2021.104152. Jiao, D., Lesage, K., et al. (2021b). “Flow behavior of cementitious-like suspension with nano-Fe3O4 particles under external magnetic field.” Materials and Structures/Materiaux et Constructions, 54(6). doi:10.1617/s11527-021-01801-y. Jiao, D., Shi, C., and Yuan, Q. (2018). “Influences of shear-mixing rate and fly ash on rheological behavior of cement pastes under continuous mixing.” Construction and Building Materials, 188, 170–177. doi:10.1016/j.conbuildmat.2018.08.091. Kashani, A., et al. (2014). “Modelling the yield stress of ternary cement-slag-fly ash pastes based on particle size distribution.” Powder Technology, 266, 203–209. doi:10.1016/j.powtec.2014.06.041. Khayat, K. H. (1995). “Effects of antiwashout admixtures on fresh concrete properties.” Materials Journal, 92(2), 164–171. Kim, J. H., et al. (2017). “Rheology of cement paste under high pressure.” Cement and Concrete Composites, 77, 60–67. doi:10.1016/j.cemconcomp.2016.11.007. Koehler, E. P., and Fowler, D. W. (2004). Development of a portable rheometer for fresh Portland cement concrete, Technical Reports, International Center for Aggregates Research, The University of Texas at Austin, Austin, United States. Kong, X., Zhang, Y., and Hou, S. (2013). “Study on the rheological properties of Portland cement pastes with polycarboxylate superplasticizers.” Rheologica Acta, 52(7), 707–718. doi:10.1007/ s00397-013-0713-7. Kwan, A. K. H., and Fung, W. W. S. (2013) ‘Effects of SP on flowability and cohesiveness of cementsand mortar’, Construction and Building Materials, 48, 1050–1057. Laskar, A. I., and Talukdar, S. (2008). “Rheological behavior of high performance concrete with mineral admixtures and their blending.” Construction and Building Materials, 22(12), 2345– 2354. doi:10.1016/j.conbuildmat.2007.10.004. Lee, S. H., et al. (2003). “Effect of particle size distribution of fly ash–cement system on the fluidity of cement pastes.” Cement and Concrete Research, 33(5), 763–768. Leemann, A., and Winnefeld, F. (2007). “The effect of viscosity modifying agents on mortar and concrete.” Cement and Concrete Composites, 29(5), 341–349. Li, G., and Wu, X. (2005). “Influence of fly ash and its mean particle size on certain engineering properties of cement composite mortars.” Cement and Concrete Research, 35(6), 1128–1134. Li, Y., and Kwan, A. K. H. (2014). “Ternary blending of cement with fly ash microsphere and condensed silica fume to improve the performance of mortar.” Cement and Concrete Composites, 49, 26–35. Liu, J., et al. (2017). “Influence of superplasticizer dosage on the viscosity of cement paste with low water-binder ratio.” Construction and Building Materials, 149, pp. 359–366. doi:10.1016/j. conbuildmat.2017.05.145. Lowke, D., et al. (2010). “Effect of cement on superplasticizer adsorption, yield stress, thixotropy and segregation resistance”, in Design, production and placement of self-consolidating concrete, edited by Khayat, K. H. and Feys, D. Springer, Montreal, Canada, pp. 91–101. Lowke, D. (2018). “Thixotropy of SCC—A model describing the effect of particle packing and superplasticizer adsorption on thixotropic structural build-up of the mortar phase based on interparticle interactions.” Cement and Concrete Research, 104, 94–104. doi:10.1016/j. cemconres.2017.11.004. Lu, C., Yang, H., and Mei, G. (2015). “Relationship between slump flow and rheological properties of self compacting concrete with silica fume and its permeability.” Construction and Building Materials, 75, 157–162. Ma, K., et al. (2013a). “Rheological properties of compound pastes with cement-fly ash-limestone powder.” Journal of the Chinese Ceramic Society, 41(5), 582–587.

Rheology for cement paste  47 Ma, K., et al. (2013b). “Factors on affecting plastic viscosity of cement-fly ash-limestone compound pastes.” Journal of the Chinese Ceramic Society, 41(11), 1481–1486. Ma, S., Qian, Y., and Kawashima, S. (2018). “Experimental and modeling study on the non-linear structural build-up of fresh cement pastes incorporating viscosity modifying admixtures.” Cement and Concrete Research, 108(January), 1–9. doi:10.1016/j.cemconres.2018.02.022. Malhotra, V. M., and Mehta, P. K. (2004). Pozzolanic and cementitious materials. CRC Press, Boca Raton, FL. Mardani-Aghabaglou, A., et al. (2013). “Effect of different types of superplasticizer on fresh, rheological and strength properties of self-consolidating concrete.” Construction and Building Materials, 47, 1020–1025. Mostafa, A. M., Diederich, P., and Yahia, A. (2015). “Effectiveness of rotational shear in dispersing concentrated cement suspensions.” Journal of Sustainable Cement-Based Materials, 4(3–4), 205–214. Nanthagopalan, P., et al. (2008). “Investigation on the influence of granular packing on the flow properties of cementitious suspensions.” Cement and Concrete Composites, 30(9), 763–768. Nawa, T., Ichiboji, H., and Kinoshita, M. (2000). “Influence of temperature on fluidity of cement paste containing superplasticizer with polyethylene oxide graft chains.” ACI Special Publication, pp. 195–210. doi:10.14359/9912. Nehdi, M., and Al Martini, S. (2009). “Estimating time and temperature dependent yield stress of cement paste using oscillatory rheology and genetic algorithms.” Cement and Concrete Research, 39(11), 1007–1016. doi:10.1016/j.cemconres.2009.07.011. Nehdi, M., Mindess, S., and Aıtcin, P. C. (1998). “Rheology of high-performance concrete: Effect of ultrafine particles.” Cement and Concrete Research, 28(5), 687–697. Overbeek, J. T. G. (1984). “Interparticle forces in colloid science.” Powder technology, 37(1), 195–208. Pal, S. C., Mukherjee, A., and Pathak, S. R. (2003). “Investigation of hydraulic activity of ground granulated blast furnace slag in concrete.” Cement and Concrete Research, 33(9), 1481–1486. Papanastasiou, T. C. (1987). “Flows of materials with yield.” Journal of Rheology, 31(5), 385–404. Papo, A., and Piani, L. (2004). “Effect of various superplasticizers on the rheological properties of Portland cement pastes.” Cement and Concrete Research, 34(11), 2097–2101. doi:10.1016/j. cemconres.2004.03.017. Park, C. K., Noh, M. H., and Park, T. H. (2005). “Rheological properties of cementitious materials containing mineral admixtures.” Cement and Concrete Research, 35(5), 842–849. Perrin, J. (1916). Atoms (translated by Hammick D. LL.). D. Van Nostrand Company, New York. Perrot, A., et al. (2012). “Yield stress and bleeding of fresh cement pastes.” Cement and Concrete Research, 42(7), 937–944. doi:10.1016/j.cemconres.2012.03.015. Petit, J. Y., et al. (2007). “Yield stress and viscosity equations for mortars and self-consolidating concrete.” Cement and Concrete Research, 37(5), 655–670. doi:10.1016/j.cemconres.2007.02.009. Petit, J. Y., Khayat, K. H., and Wirquin, E. (2006). “Coupled effect of time and temperature on variations of yield value of highly flowable mortar.” Cement and Concrete Research, 36(5), 832–841. doi:10.1016/j.cemconres.2005.11.001. Petit, J. Y., Khayat, K. H., and Wirquin, E. (2009). “Coupled effect of time and temperature on variations of plastic viscosity of highly flowable mortar.” Cement and Concrete Research, 39(3), 165–170. doi:10.1016/j.cemconres.2008.12.007. Prakash, N., and Santhanam, M. (2006). “A study of the interaction between viscosity modifying agent and high range water reducer in self compacting concrete”, in Measuring, monitoring and modeling concrete properties, edited by Konsta-Gdoutos, M. S. Springer, Dordrecht, Netherlands, pp. 449–454. Rahman, M. A., and Nehdi, M. (2003). “Effect of geometry, gap, and surface friction of test accessory on measured rheological properties of cement paste.” Materials Journal, 100(4), 331–339. Rahman, M. K., Baluch, M. H., and Malik, M. A. (2014). “Thixotropic behavior of self compacting concrete with different mineral admixtures.” Construction and Building Materials, 50, 710–717. Roussel, N. (2006). “A thixotropy model for fresh fluid concretes: Theory, validation and applications.” Cement and Concrete Research, 36(10), 1797–1806. doi:10.1016/J.CEMCONRES.2006.05.025. Roussel, N., et al. (2010). “Steady state flow of cement suspensions: A micromechanical state of the art.” Cement and Concrete Research, 40(1), 77–84. doi:10.1016/j.cemconres.2009.08.026.

48  Rheology of Fresh Cement-Based Materials Schmidt, W., et al. (2010). “Effects of superplasticizer and viscosity-modifying agent on fresh concrete performance of SCC at varied ambient temperatures”, in Design, Production and Placement of Self-Consolidating Concrete, edited by Khayat, K., and Feys, D. Springer, Dordrecht, pp. 65–77. Secrieru, E., et al. (2018). “Changes in concrete properties during pumping and formation of lubricating material under pressure.” Cement and Concrete Research, 108, 129–139. doi:10.1016/j. cemconres.2018.03.018. Shen, W., et al. (2021). “Change in fresh properties of high-strength concrete due to pumping.” Construction and Building Materials, 300, 1–32. doi:10.1016/j.conbuildmat.2021.124069. Shi, Y., Matsui, I., and Guo, Y. (2004). ‘A study on the effect of fine mineral powders with distinct vitreous contents on the fluidity and rheological properties of concrete.” Cement and Concrete Research, 34(8), 1381–1387. Struble, L. J., and Jiang, Q. (2005). “Effects of air entrainment on rheology.” ACI Materials Journal, 101, 448–456. Tang, X., et al. (2014). “Correlation between slump flow and rheological parameters of compound pastes with high volume of ground slag.” Journal of the Chinese Ceramic Society, 42(5), pp. 648–652. Tattersall, G., and Banfill, P. (1983). The rheology of fresh concrete. Pitman Books Limited, London, England. Tattersall, G. H. (1991). Workability and quality control of concrete. CRC Press, Boca Raton, FL. Taylor, H. F. W. (1997). Cement chemistry. Thomas Telford, London. Toledano-Prados, M., et al. (2013). “Effect of polycarboxylate superplasticizers on large amounts of fly ash cements.” Construction and Building Materials, 48, 628–635. Tregger, N. A., Pakula, M. E., and Shah, S. P. (2010). “Influence of clays on the rheology of cement pastes.” Cement and Concrete Research, 40(3), 384–391. doi:10.1016/j.cemconres.2009.11.001. Uysal, M., and Yilmaz, K. (2011). “Effect of mineral admixtures on properties of self-compacting concrete.” Cement and Concrete Composites, 33(7), 771–776. doi:10.1016/j. cemconcomp.2011.04.005. Vance, K., et al. (2013). “The rheological properties of ternary binders containing Portland cement, limestone, and metakaolin or fly ash.” Cement and Concrete Research, 52, 196–207. Wallevik, J. E. (2003). Rheology of particle suspensions: Fresh concrete, mortar and cement paste with various types of lignosulfonates. Fakultet for ingeniørvitenskap og teknologi. Wallevik, J. E., and Wallevik, O. H. (2017). “Analysis of shear rate inside a concrete truck mixer.” Cement and Concrete Research, 95, 9–17. doi:10.1016/j.cemconres.2017.02.007. Wallevik, O. H., and Wallevik, J. E. (2011). “Rheology as a tool in concrete science: The use of rheographs and workability boxes.” Cement and Concrete Research, 41(12), 1279–1288. Williams, D. A., Saak, A. W., and Jennings, H. M. (1999). “Influence of mixing on the rheology of fresh cement paste.” Cement and Concrete Research, 29(9), 1491–1496. doi:10.1016/ S0008-8846(99)00124-6. Yahia, A. (1999). “Effect of limestone powder on rheological behavior of highly-flowable mortar”. Proceeding of the Japan Concrete Institute, 21(2), 559–564. Yamada, K., et al. (1999). “Influence of temperature on the dispersibility of polycarboxylate type superplasticizer for highly fluid concrete”, rilem.net. Available at: https://www.rilem.net/gene/ main.php?base=500218&id_publication=12&id_papier=1310 (Accessed: 3 April 2018). Yamada, K., Hanehara, S., and Matsuhisa, M. (1998). “Fluidizing mechanism of cement paste added with polycarboxylate type superplasticizer analyzed from the point of adsorption behavior.” Proceedings of JCI, 20(2), 63–78. Yamada, K., Ogawa, S., and Hanehara, S. (2000). “Working mechanism of poly-beta-naphthalene sulfonate and polycarboxylate superplasticizer types from point of cement paste characteristics.” Special Publication, 195, 351–366. Yamada, K., Ogawa, S., and Hanehara, S. (2001). “Controlling of the adsorption and dispersing force of polycarboxylate-type superplasticizer by sulfate ion concentration in aqueous phase.” Cement and Concrete Research, 31(3), 375–383. doi:10.1016/S0008-8846(00)00503-2.

Rheology for cement paste  49 Yang, M., Neubauer, C. M., and Jennings, H. M. (1997). “Interparticle potential and sedimentation behavior of cement suspensions: Review and results from paste.” Advanced Cement Based Materials, 5(1), 1–7. Yoshioka, K., et al. (1997). “Role of steric hindrance in the performance of superplasticizers for concrete.” Journal of the American Ceramic Society, 80(10), 2667–2671. Yun, K.-K., Choi, S.-Y., and Yeon, J. H. (2015). “Effects of admixtures on the rheological properties of high-performance wet-mix shotcrete mixtures.” Construction and Building Materials, 78, 194–202. Zhang, X., and Han, J. (2000). “The effect of ultra-fine admixture on the rheological property of cement paste.” Cement and Concrete Research, 30(5), 827–830. doi:10.1016/S0008-8846(00)00236-2. Zhang, Y., and Zhang, X. (2002). “Relationship between the content of slag powder and the characteristics of particle group and rheological property of cement paste.” Bulletin of the Chinese Ceramic Society, 21(6), 63–67, 75. Zhou, Z., et al. (1999). “The yield stress of concentrated flocculated suspensions of size distributed particles.” Journal of Rheology, 43(3), 651–671. Zuo, Y. F., Sui, T. B., and Wang, D. M. (2004). “Effect of superplasticizers on rheologic performance of fresh cement paste.” Hunningtu (Concrete) (09), 38–39.

Chapter 3

Rheological properties of fresh concrete materials

3.1 GENERAL CONSIDERATIONS FOR GRANULAR MATERIALS The term “the rheology of mortar and concrete” mainly indicates the evolution of viscosity, plasticity, and elasticity under shear stress (Barnes et al., 1989). Yield stress and plastic viscosity as two fundamental physical parameters are of great importance in cement-based materials. On the one hand, the rheological parameters are effective for characterization of the workability, prediction of the flow behavior, and evaluations of the pumpability, stability, and formwork filling of mortar and concrete (Kwon et al., 2013, Roussel 2007, Roussel et al., 2010). The rheological properties also play significant roles in describing the chemical hydration, microstructural formation, and setting process of fresh cementitious materials (Bogner et al., 2020, Iqbal Khan et al., 2016, Jiao et al., 2019a, Sant et al., 2008). On the other hand, the adjustment of rheological parameters such as static yield stress and plastic viscosity can be helpful to strike a balance between pumpability, extrudability, and ­printability (De Schutter et al., 2018, Roussel 2018, Yuan et al., 2019). Consequently, the rheological parameters are vital for the preparation of high-performance concrete (Jiao et al., 2018, Wallevik, 2003). Fresh cement-based materials such as mortar and concrete can be regarded as highly concentrated suspensions with aggregate particles suspending in a cement paste phase. The particle size in a cementitious system varies from several nanometers to dozens of millimeters. Therefore, fresh cementitious materials can be viewed as a two-phase system, i.e., fluid-paste and solid-phase aggregates (Jiao et al., 2017b, Nielsen, 2001). Depending on the content of the cement utilized, fresh concrete is divided into three categories, i.e., lean concrete with cement content lower than 10%, normal concrete with cement content ranging from 10% to 15%, and rich concrete with cement content higher than 15% (Talbot et al., 1923). In the present chapter, the theoretical correlations between paste volume (or aggregate volume fraction) and rheological parameters for different types of concrete are introduced. The effect of characteristics of aggregates such as particle size, shape, and roughness on the rheological properties is discussed from theoretical and experimental perspectives. Furthermore, the influences of external factors such as mixing procedure, shear history, and measuring geometry on the rheology of fresh concrete are briefly illustrated. 3.2 FLOW REGIMES OF CONCRETE Aggregate accounts for more than 75% of the total volume of conventional vibrated concrete and 60% of self-compacting concrete (SCC) (Jiao et al., 2017b). From the rheological point of view, aggregates with large grain size and rough surface restrict the flow of concrete and increase the yield stress and plastic viscosity of concrete than that of cement DOI: 10.1201/9781003265313-3

51

52  Rheology of Fresh Cement-Based Materials

paste. Generally, the chemical compositions of aggregate have little effect on the rheological properties (Mahaut et al., 2008b). In this section, the theoretical calculations and empirical relationships between rheological properties (plastic viscosity and yield stress) and aggregate volume fraction are illustrated. The excess paste theory is also highlighted.

3.2.1 Relationships between aggregate volume fraction and concrete rheology 3.2.1.1 Viscosity vs aggregate volume fraction The derivation of viscosity of a concentrated suspension is originated from the well-known Einstein viscosity equation, as expressed in Eq. (3.1):

ηr = 1 + [η ]φ (3.1)

where ηr is the relative viscosity of a suspension to its suspending fluid, ϕ is the volume concentration of particles, and [η] is the intrinsic viscosity of solids, which is a measure of the effect of individual particles on the viscosity. The value of intrinsic viscosity is 2.5 for spherical particles, 3–5 for angular but presumably equant particles, and 4–10 for rods or fibers (Struble and Sun, 1995). The Einstein equation is restricted to dilute suspensions with very low particle volume fractions (lower than 0.05), supposing no particle interaction (Quemada, 1984). Subsequently, Robinson proposed a modified Einstein model for higher concentrations, by considering that the viscosity depends on both the volume fraction of particles and the volume of free liquid in the suspension (Robinson, 1949), as expressed by:

ηr = 1 +

kφ (3.2) 1 − Vφ

where k is a constant, equal to the intrinsic viscosity of particles (Ren, 2021), and V is the bulk volume per unit volume of solid, regarding as the reciprocal of maximum packing fraction which will be introduced later. Besides, considering the interaction between solid particles for medium and high concentrations, Roscoe proposed a modified viscosity model (Roscoe, 1952):

ηr = (1 − 1.35φ)−[η ] (3.3)

As well known, the viscosity generally increases with the solid concentrations. However, due to the crowding effect, the suspension with a sufficiently high concentration cannot flow when solid particles become packed so tightly. This limit concentration is called maximum volume fraction ϕm , depending on the size distribution and shape of the particles. For monosized spherical particles, the maximum volume fraction is near 0.6–0.7, while it is higher for polydisperse particles (Struble and Sun, 1995). Based on the Einstein viscosity equation, Mooney stated a functional equation considering the maximum packing volume fraction of rigid spheres (Mooney, 1951), which can be expressed as follows:



  [η ]φ ηr = exp  1 − φ  φm

   (3.4)  

Rheological properties of fresh concrete materials  53

The Mooney equation can be used to describe the viscosity of suspensions with low particle volume fractions (Struble and Sun, 1995). For the suspensions with high concentrations, the Mooney equation is no longer applicable. In this case, the Krieger–Dougherty equation, as shown in Eq. (3.5), was proposed to describe the full range of concentrations (Krieger and Dougherty, 1959):

 φ  ηr =  1 − φm  

−[η ]φm

(3.5)

where ϕm is the maximum packing fraction, which is very sensitive to particle size distribution and particle shape. It should be mentioned that the value of [η]ϕm is always around 2 for various suspensions. In this case, the Krieger–Dougherty model can be simplified to the Maron–Piece model, as stated by (Ren, 2021): −2



 φ  ηr =  1 − (3.6) φm  

Back again the Krieger–Dougherty model, if considering fresh concrete as a suspension with multi-size particles and ignoring the interaction between particles of different sizes, the viscosity of the multimodal suspensions can be derived from the unimodal viscosity of each size, namely the Farris model (Farris, 1968):



 φ  η = ηS  1 − 11  φm  

1 −[η1 ]φm

 φ2   1 − φ 2  m

2 −[η2 ]φm

(3.7)

Typically, solid particles in fresh concrete are divided into coarse and fine aggregates. In this case, the Farris model can be modified in the following version (Noor and Uomoto, 2004):

 S  ηC = ηP  1 −  Slim 

−[ηFA ]Slim

 G   1 − G  lim

−[ηCA ]Glim

(3.8)

where S and S lim are the sand volume fraction and its maximum solid volume, respectively. G and Glim are the gravel volume fraction and the maximum gravel solid volume. The S lim and Glim are selected as 0.643 and 0.575 in their study, respectively. ηC and ηP are the viscosity of concrete and paste (Pa.s), respectively. [η FA] and [η CA] are the intrinsic viscosity of fine and coarse aggregates with values of 1.9 and 3.2 in their study, respectively. Experimental results indicate that the multimode Farris model is reasonable and applicable for determining the viscosity of fresh mortar and concrete (Noor and Uomoto, 2004). 3.2.1.2 Yield stress vs aggregate volume fraction Considering the fact that aggregates distribute in cement paste matrix uniformly, the shear stress of mortar and/or concrete is assumed to be the sum of the shear stresses resulting from the yield stress and flow of cement paste, the shear stress induced by the aggregate particle movement, and the interaction between cement paste and aggregates (Lu et al., 2008, Wang et al., 2020). Therefore, the aggregate volume fraction significantly affects the yield stress of the mortar and/or concrete.

54  Rheology of Fresh Cement-Based Materials

If aggregates are divided into smaller subclasses i and the interactions of these subclasses are considered, a powerful semi-empirical model was established to correlate the yield stress of concrete with the volume fraction (De Larrard and Sedran, 2002), as shown in Eq. (3.9):

τ 0,C = 2.537 +



∑ (0.736 − 0.216 log (d )) K ′ +  0.224 + 0.910  1 − PP  i

i

i





3

  KC′ (3.9) 

φ ; ϕi, ϕi*, and di are the volume frac1 − φi∗ tion, the maximum packing volume fraction, and the size of particles of class i, respectively; subscript C is the cement; and P and P* are the superplasticizer dosages and saturation dosage, respectively. It should be noted that this model is only applied to the case of concrete only with one binder (Toutou and Roussel, 2006). Similarly, Noor and Uomoto stated that the yield stress of concrete is assumed to be a function of mortar and the volume fraction of aggregates (Noor and Uomoto, 2004), which can be expressed as follows: where τ0,C is the yield stress of concrete (Pa); Ki′ =



τ 0,C = τ 0,m + f (φ) (3.10)

where τ0,C and τ0,m are the yield stress of concrete and mortar (Pa), respectively. ϕ is the total aggregate volume fraction. It should be mentioned that the model in Eq. (3.10) based on experimental data of mortar and concrete is completely empirical. Considering fresh concrete as a concentrated suspension with aggregates (i.e., sand and gravel) suspending in cement paste, the yield stress of the suspension is proportional to that of the constitutive cement paste (Yammine et al., 2008). The general form of the relation is expressed as follows:

τ 0,C ≈ τ 0,cp f (φ φm ) (3.11)

where τ0,C and τ0,cp are the yield stress of the concrete and the corresponding cement paste (Pa), respectively. ϕm is the maximum packing fraction. A semi-empirical calculation for the maximum packing fraction can be expressed by Eq. (3.12) (Hu and de Larrard, 1996):

φm = 1 − 0.45 ( dmin dmax )

0.19

(3.12)

where dmin and dmax are the smallest and largest grain diameters in the granular skeleton (m). Independent of the mixture proportion of cement paste, Eq. (3.11) is of great importance to study the relative yield stress of the concrete, i.e., the ratio of yield stresses between concrete and cement paste (τ0,C /τ0,cp). Based on Eq. (3.11), two examples are given here to illustrate the modified theoretical relationship between yield stress and volume fraction. The first example is about the domination theory of hydrodynamic or frictional effects depending on the interparticle distance, i.e., fluid–particle or particle–particle interactions (Yammine et al., 2008). At low volume fractions, the interparticle interactions are hydrodynamic. The relative motions of aggregates imply the flow of cement paste. During the flowing of the concrete, additional energy dissipation induced by the possible analogous effect occurs, leading to the greater yield stress of the concrete than that of the cement paste alone. Consequently, the yield stress of concrete increases with the volume fraction of aggregates. At higher volume fractions, however, direct contacts between aggregate particles may

Rheological properties of fresh concrete materials  55

Figure 3.1  Concrete yield stress as a function of aggregate volume fraction. (Adapted from Yammine et al., 2008.)

occur. Remarkably, a “true” direct contact is hardly defined due to the particle interactions, particle roughness, and hydrodynamic effects. Instead, the physical behavior of particle contacts can be associated with the existence of a continuous network of particles in contact. In other words, this phenomenon occurs when the aggregate volume fraction is larger than a critical value ϕc. According to experimental and numerical results, ϕc should be situated around 0.5 for uniform spheres (Onoda and Liniger, 1990), and it should be increased in the case of poly-dispersed systems. A representative relationship between concrete yield stress and aggregate volume fraction is presented in Figure 3.1, where the maximum packing of aggregates is equal to 0.84 and the transition volume fraction is in the order of 0.65. It can be seen that the direct contact between aggregate particles can be neglected below the transition value. In this regime, the effect of aggregate particles on the rheological behavior of the concrete is regarded as purely hydrodynamic. Inversely, direct frictional contacts between particles start to dominate the rheological behavior above this transition value, and their highly dissipative nature strongly increases the yield stress of the concrete. The transition volume fraction between the frictional and hydrodynamic regimes is independent of the selection of the multi-scale approach (Yammine et al., 2008). The second example here is the quantitative relation between relative yield stress and volume fraction of aggregates. For a suspension following the Herschel–Bulkley model, the relative yield stress of concrete compared to the corresponding cement paste only depends on the solid volume fraction (Chateau et al., 2008). The relation between the yield stress and the volume fraction follows the Chateau–Ovarlez–Trung model (Chateau et al., 2008), as expressed in Eq. (3.13) with ϕm of 0.57:

τ c (φ) 1−φ = (3.13) τ c (0) 1 − (φ / φ)2.5φm

It should be noted here that the maximum volume fraction ϕm is not selected as the so-called random dense packing fraction (ϕ RDP), which is around 0.65 for spherical monodisperse particles. Instead, the critical volume fraction where direct contacts become important, which is the application limit of models only considering hydrodynamic interactions and also the limit between self-compacting concrete (SCC) and ordinary rheology concrete (Yammine

56  Rheology of Fresh Cement-Based Materials

et al., 2008), is used as the maximum volume fraction for the sharp increase of yield stress. Therefore, the parameter of ϕm in Eq. (3.13) is also called rheology divergence packing fraction ϕdiv (Ovarlez et al., 2006). Hafid et al. (2016) stated that the rheology divergence packing fraction is related to the random dense and loose packing fractions, which can be expressed as follows:

φdiv = 0.64φRDP = 0.88φRLP (3.14)

If considering fresh concrete as a suspension with grains dispersing in mortar matrix while the fresh mortar is considered as a suspension with sand suspending in cement paste, the Chateau–Ovarlez–Trung model can be modified to the following version (Choi et al., 2013):

τ c (φ) = τ c (0)

1 − φg

1 − φs

(1 − φs

φ s ,m )

2.5φs ,m

(1 − φg

φ g ,m )

2.5φ g ,m

(3.15)

where ϕs and ϕg are the volume fraction of sand and grains, respectively, and ϕs, m and ϕg, m are the maximum volume fraction of sand and grains, respectively. Recently, Kabagire et al. (2017) stated that the constant 2.5 in Eq. (3.13) can be replaced by a fitted coefficient [η]*, representative of the modified intrinsic viscosity depending on the geometric features of particles and the shear conditions (Kabagire et al., 2019). In this case, Eq. (3.13) can be converted into:

τ c (φ) = τ c (0)

1−φ

(1 − φ φ m )

[η ]*φm

(3.16)

where the modified intrinsic viscosity [η]* is determined by nonlinear regression. Furthermore, after taking the effect of paste-to-sand volume fraction into account, the relative yield stress of SCC can be calculated by (Kabagire et al., 2019): d



 (VP/VS)e  τ c (φ) 1−φ  =  (3.17) ∗ τ c (0)  (1 − φ φm )[η ] φm 

where VP/VS is the paste-to-sand volume fraction, and d and e are constants with the value of 1.1 and 1.24 in their study, respectively (Kabagire et al., 2019). It is found that the relative yield stresses of SCC with crushed limestone coarse aggregate between measured and predicted by Eq. (3.17) show a good correlation coefficient of 0.75.

3.2.2 Excess paste theory Cement paste is an essential component of concrete. A schematic diagram of the fresh concrete model is presented in Figure 3.2. It can be seen that the cement paste in concrete can be divided into two parts: one is the paste filling the voids between aggregates, and the other is the paste coating aggregates. The second part is defined as the excess paste. The excess paste layer is generally used to lubricate the solid grains and provide the flowability of the concrete mixture (Jiao et al., 2017b).

Rheological properties of fresh concrete materials  57

Figure 3.2  Schematic diagram of fresh concrete (SCC) model. (Adapted from Reinhardt and Wüstholz, 2006.)

There are several approaches to calculate the excess paste thickness. A rough approximation of the average interparticle distance can be expressed as follows (Yammine et al., 2008):

(

b = −d 1 − (φ φm )

−1/3

) (3.18)

where b is the average interparticle distance (m), d is the particle size (m), ϕ is the particle volume fraction, and ϕm is the maximum packing fraction. It should be mentioned that the average distance is equal to two times the excess paste thickness. A simple approach to calculating the volume of excess paste is described by Reinhardt and Wüstholz (2006) as follows:

Vpaste,ex = Vpaste − VA,void = Vpaste −

 mA  ρ A − 1 (3.19)  ρ A  ρ A,bulk 

where Vpaste, ex is the volume of excess paste (m3); Vpaste is the total paste volume (m3); VA, void is the volume of voids between aggregates (m3); mA and ρA are the mass and density of ­aggregates (kg), respectively; and ρA, bulk is the loose bulk density of aggregates (kg/m3). If a spherical paste layer on the aggregates is assumed and the excess paste thickness is independent of the particle size, the excess paste thickness can be calculated using Eq. (3.20):

Vpaste,ex =

4 π 3

∑ n (( r + t i

i

i

)

3

paste,ex

)

− ri3 (3.20)

58  Rheology of Fresh Cement-Based Materials

Figure 3.3 A representative relationship between rheological properties and excess paste thickness at the water-to-cement ratio of 0.4, the sand-to-aggregate ratio of 0.42, and the superplasticizer dosage of 1%. (Adapted from Jiao et al., 2017a.)

where tpaste, ex is the excess paste thickness (m), and ri and ni are the radius (m) and the number of a particle of class i, respectively. Another calculation method of excess paste thickness is as follows (Oh et al., 1999):

V  10  t paste,ex =  1 − 100 S  (3.21)  CS  SSVS

where tpaste, ex is the excess paste thickness (mm), VS is the aggregate-to-mortar volume ratio, C S is the sand volume divided by its bulk volume (%), and S S is the specific surface area of aggregates (cm 2 /cm3). Based on the concept of excess paste thickness, the amount of excess paste plays a dominant role in determining the rheological properties of mortar or concrete than the total amount of paste. Furthermore, the influencing factors on concrete rheology such as paste volume and aggregate volume can be transferred into the parameter of excess paste thickness (Reinhardt and Wüstholz, 2006, Ren et al., 2021). A typical relationship between rheological parameters and excess paste thickness for conventional concrete is shown in Figure 3.3. At relatively low excess paste thickness, the concrete shows higher yield stress and plastic viscosity, and lower flowability. In this case, the yield stress is relatively more affected by the excess paste thickness than the plastic viscosity. With the increase of excess paste thickness, the yield stress slightly increases, while the plastic viscosity gradually decreases. The flowability and stability of the concrete are improved. At relatively high excess paste thickness, the concrete shows very high slump value, but lower yield stress and plastic viscosity. In this case, the stability of the concrete is not good enough, and a slight segregation phenomenon can be observed (Jiao et al., 2017a). The yield stress of mortar and concrete can also be quantitatively predicted by the paste layer thickness (Lee et al., 2018). Assuming that the paste layer thickness is only dependent on the properties of cement paste, the layer thickness can be regarded as a constant value regardless of the particle volume fraction. Based on this assumption, the yield stress of SCC concerning the suspending fluid can be estimated by Eq. (3.22):

Rheological properties of fresh concrete materials  59 n



3 3    τc b b  = k  ϕ1  1 +  + ϕ 2  1 +  − ϕ c  (3.22)  τ0 d1  d2    

where τc and τ0 are the yield stress of concrete and the corresponding cement paste (Pa), respectively; b is the paste layer thickness (mm); k is a constant; and d1 and d2 are the particle size of fine and coarse aggregates (mm), respectively. φ1 and φ2 are the volume fraction of fine and coarse aggregates, respectively. φt is the percolation threshold, with the value of 0.29 for mono-sized suspension. With the experimental results of corresponding mortar with 90%-volume-fraction aggregates and wet-sieved mortar, the yield stress of concrete can be acceptably predicted by this model (Lee et al., 2018). 3.3 INFLUENCE OF AGGREGATE CHARACTERISTICS

3.3.1 Aggregate volume fraction From the experimental point of view, a typical relationship between plastic viscosity of mortar/concrete and aggregate volume fraction is presented in Figure 3.4. It can be clearly observed that aggregates only play a significant role at high solid volume fractions. In addition, for the cement-based materials with low aggregate volume fraction, the effect of aggregate size, gradation, and surface texture is not significant (Li and Liu, 2021, Lu et al., 2008). Based on the assumption that plastic viscosity is the microscopic indication of the flow behavior of water in the voids between solid particles (De Larrard and Sedran, 2002), the results indicate that the contribution of aggregate volume fractions on the plastic viscosity of fresh lean concrete is possibly much higher than that of rich concrete. This can be explained by the increased collisions and frictions between solid particles at high aggregate volume fractions (Okamura and Ouchi, 2003).

Figure 3.4 Relationship between plastic viscosity and relative particle concentration. SP is superplasticizer, and SF is silica fume. (Adapted from De Larrard and Sedran, 2002.)

60  Rheology of Fresh Cement-Based Materials

Figure 3.5 Dimensionless yield stress vs volume fraction of beads with various particle sizes in various suspensions. The solid line is based on Eq. (3.13) with ϕm of 0.57. (Adapted from Mahaut et al., 2008a.)

In the case of yield stress, through investigating the suspensions with rigid noncolloidal particles embedded in a yield stress fluid, Mahaut et al. (2008b) found that the relative yield stress of concrete to its cement paste only depends on the yield stress value of suspending cement paste and volume fraction of particles. However, the relative yield stress is independent of the physicochemical properties of the matrix, the bead material, and the particle size. A typical relationship between the relative yield stress and the solid volume fraction is shown in Figure 3.5. It can be observed that the relative yield stress of suspensions shows an exponential increase with the increase of particle volume fraction, with a limited increase degree at volume fraction lower than 30% and a sharp increase at volume fraction approach to 50% (Mahaut et al., 2008b).

3.3.2 Gradation and particle size According to the excess paste theory, it can be easily concluded that for a concrete mixture with a fixed content of cement paste, the higher the aggregate packing density, the lower the fraction of voids, and thus higher the content of excess paste acting as lubrication effect. Therefore, the properties of concrete are strongly affected by the packing of aggregate particles. The existing particle packing models for mono-sized particles and poly-dispersed systems were summarized by Kumar and Santhanam (2003) and Roussel (2011). In the context of skeleton for cement-based materials, the aggregate gradation optimization process is to achieve a good aggregate packing by selecting the proper sizes and proportions of small particles to fill larger voids approaching an ideal gradation curve. An optimal aggregate gradation provides a concrete mixture with excellent properties. In this part, the most useful continuous models in cementitious materials are briefly illustrated, and the effects of particle size and gradation on the rheological properties of concrete are discussed as well. The most well-known and applicable ideal curve of particle packing is the theoretical particle size distribution described by the Fuller–Thompson model (Fuller and Thompson, 1907), which can be expressed as follows:

P = ( di Dmax ) (3.23) q

where P is the cumulative passing, di is the particle diameters under consideration (m), Dmax is the nominal maximum particle size diameter (m), and q is the packing exponent of 0.5 for Fuller and Thompson (1907), and then revised to 0.45 by Talbot et al. (1923). It should be noted that the Fuller and Thompson curve is only valid for the aggregate skeleton, and it gives good results if stiff concrete mixes with low workability are used.

Rheological properties of fresh concrete materials  61

Figure 3.6 Influence of the size of fine aggregate on yield stress and plastic viscosity of mortar. (Adapted from Han et al., 2017.). G indicates mono-sized grains, and M represents mixed sands.

Considering the aggregates and cementitious materials, i.e., all the solid grains in the mixture, a modified Andreasen & Andersen model proposed by Funk and Dinger (2013) can be applied:

P=

q diq − dmin (3.24) q q − dmin dmax

where P is the cumulative passing; di is the particle diameters under consideration (m); Dmin and Dmax are the minimum and nominal maximum particle size diameters (m), respectively; and q is the packing modulus. The results of Mueller et al. (2014) showed that the modified Andreasen & Andersen model with q of about 0.27 can be used to express the particle size distribution of all the solids in SCC with low powder content. As a result, this model is widely used in the mixture design of conventional vibrating concrete and even SCC. From the viewpoint of experimental results, smaller aggregate particle size generally results in higher yield stress and plastic viscosity of cement-based materials, due to the high specific surface area and thus high water demand (Jiao et al., 2017b). At low aggregate volume fractions, the contribution of particle size to the rheological properties is not significant, and the influence can only be obvious at higher volume fractions (Lu et al., 2008). Hu (2005) conducted a series of experiments investigating the effect of aggregate on the rheology of cement-based materials. For mortars with a fixed water-to-cement (w/c) ratio of 0.5 and sand-to-cement ratio of 2, increasing particle size from 0.15 to 2.36 mm reduces the yield stress from 130 to 70 Pa, and decreases the plastic viscosity from 2 to 1.2 Pa.s. Han et al. (2017) also stated that increasing the fine aggregate size decreases the yield stress and plastic viscosity of fresh mortar. In addition, they also concluded that changing the size of fine aggregate when exceeding 0.70 mm will show no significant influence on the yield stress, as can be observed from Figure 3.6. From the viewpoint of concrete, the particle size of coarse aggregate plays a more significant role in the yield stress compared to the plastic viscosity (Hu, 2005). As for the gradation, graded aggregates generally have lower uncompacted void content than single-sized aggregates (Hu and Wang, 2007, Hu and Wang, 2011). Consequently, more excess mortar is generated to lubricate the particles for graded aggregates, and thus the yield stress and plastic viscosity of concrete mixtures with graded coarse aggregates are significantly lower than

62  Rheology of Fresh Cement-Based Materials

that with single-sized coarse aggregates. Santos et al. (2015) stated that concrete mixtures with continuous skeleton showed higher slump flow values, whereas due to the interlocking effect, the discontinuous distribution mixtures exhibited lower flowability. In addition, the poor grading of fine aggregates can to an extent be corrected by finer sand with low fineness modulus. Lower volumetric replacement of crushed sand by dune sand, for example, decreases the yield stress and plastic viscosity of concrete, but higher volumetric replacements will exhibit an opposite influence because of the high specific surface area of dune sand (Bouziani et al., 2012, Park et al., 2018). The content of fines with a diameter less than 0.315 mm in coarse aggregate also has a significant influence on the rheological properties of fresh concrete (Aïssoun et al., 2015). It was found that increasing the content of fines from 8% to 18% could reduce the yield stress, increase the plastic viscosity, and improve the settlement resistance.

3.3.3 Particle morphology The morphology of aggregate particles can be quantitatively assessed by shape factors with dimensionless quantities. The shape factors represent the degree of deviation of particles from an ideal shape. Common shape factors used to characterize the morphology of aggregate particles include aspect ratio (A R), circularity (fcirc), and convexity (fconv). A schematic diagram to define the morphological parameters is shown in Figure 3.7, where the particle area is noted as S and its perimeter is noted as P. The convex perimeter (Pconv) is defined as the smallest perimeter of a polygon capturing the projected particle, presented by the gray in Figure 3.7. The aspect ratio, which is the most commonly used shape factor, is computed as the ratio of the largest (Dmax /m) and smallest (Dmin /m) dimensions of the projected particle, as expressed by Eq. (3.25):

AR =

Dmax (3.25) Dmin

Figure 3.7 Schematic diagram to define morphological parameters. (Adapted from Hafid et al., 2016.)

Rheological properties of fresh concrete materials  63

Another very common shape factor is the circularity, which is defined as follows:

4πS (3.26) P2

fcirc =

The convexity describes the surface properties of a projected particle, which can be calculated by the ratio between convex perimeter and real perimeter, which is expressed as follows:

fconv =

Pconv (3.27) P

Furthermore, the shape of aggregate particles can also be evaluated by sphericity (SP) and roundness (R) (Cordeiro et al., 2016), which can be expressed by Eqs. (3.28) and (3.29), respectively:

SP =



R=

2P0.5 (3.28) π 0.5 Dmax

4P (3.29) 2 πDmax

Generally, the aspect ratio describes the overall shape of a particle, whereas circularity captures the deviation of a particle from the perfect circle. The convexity is usually to capture the surface roughness of a particle. Several examples of geometrical shapes and their morphological parameters are listed in Table 3.1. The convexity and aspect ratio are supposed to be uncorrelated, whereas the circularity varies with both overall shape and surface roughness. Furthermore, a typical relationship between aspect ratio and circularity, convexity,

Table 3.1  Typical examples of geometrical shapes and their morphological parameters (Hafid, et al., 2016) AR

fcirc

fconv

Circle

100%

100%

100%

Square

100%

79%

100%

Rectangle

300%

59%

100%

Ellipsoid

300%

60%

100%

Star

100%

38%

79%

Shape

Illustration

Note: Underlined value means that this parameter is constant for the specific shape.

64  Rheology of Fresh Cement-Based Materials

Figure 3.8  Correlations between aspect ratio and circularity. (Adapted from Hafid et al., 2016.)

Figure 3.9  Correlations between convexity and circularity. (Adapted from Hafid et al., 2016.)

and circularity for 120 sand particles is shown in Figures 3.8 and 3.9, respectively. It can be found that convexity and circularity are well correlated for the tested particles. Therefore, aspect ratio and convexity can be used as independent morphological parameters to describe the shape of aggregate particles (Hafid et al., 2016). Moreover, the authors stated that both random dense and loose packing fractions of sand particles are strongly correlated to the aspect ratio, whereas the particle roughness seems to play a role only when the frictional contacts between particles start to dominate. Considering the volume fraction and particle shape, Geiker et al. (2002b) used the Nielsen model (Nielsen, 2001) to estimate the relative rheological parameters of concrete. Take the plastic viscosity (μC) as an example, assuming particles being ellipsoids, it can be expressed by Eq. (3.30):

µC 1 + αϕ = (3.30) µ0 1−ϕ

where μ 0 is the plastic viscosity of the corresponding cement paste (Pa.s), φ is the particle volume fraction, and α is a so-called shape function, depending on the shape functions of suspending fluid and particles. However, due to the combined effect of angularity and surface texture of aggregates, the proposed model cannot fully fit the experimental data (Geiker et al., 2002b).

Rheological properties of fresh concrete materials  65

Figure 3.10  Relationship between (a) relative yield stress, (b) relative Herschel–Bulkley consistency, and solid volume fraction of suspensions with various particle shapes. Lines are fitted by the Chateau model. (Adapted from Hafid et al., 2016.)

Figure 3.11  Relationship between rheology divergence packing fraction and the aspect ratio of particles. Aspect ratio of 100% indicates spherical particles. (Adapted from Hafid et al., 2016.)

Existing experimental results show that the rheological properties are affected by the shape and surface texture of aggregate particles. A representative relationship between the relative yield stress, the relative Herschel–Bulkley consistency, and the solid volume fraction for various particle shapes is presented in Figure 3.10. It can be seen that both the relative yield stress and the relative Herschel–Bulkley consistency increase with increasing solid volume fraction, with the divergence packing fraction dependent on the shape of the solids. The effect of particle shape on the relative consistency seems to be higher than that on the relative yield stress. Besides, the divergence packing fraction gradually decreases when the particle tends to be more irregular. A typical correlation between the measured divergence packing fraction and the aspect ratio for monodisperse particles is presented in Figure 3.11. It can be seen that the divergence packing fraction can decrease up to 40% for very elongated particles. At a given solid volume fraction, spherical particles generally result in reduced yield stress and plastic viscosity of cement-based materials due to the low uncompacted void content

66  Rheology of Fresh Cement-Based Materials

Figure 3.12  Effect of aggregate shape and sand content on rheological parameters. (Adapted from Wallevik and Wallevik, 2011.)

and low friction between particles. By contrast, irregular and elongated aggregates dramatically increase the yield stress and plastic viscosity of mortar and concrete. This can be clearly observed in Figure 3.12. It should be mentioned that the particle shape plays a more significant role in controlling the plastic viscosity and the consistency index than the yield stress (Hafid et al., 2016, Westerholm et al., 2008), and the negative effects of shaped aggregates can be reduced by increasing the paste volume (Westerholm et al., 2008). Moreover, the particle shape of finer particles (e.g., 0.125–2 mm) and the properties of filler particles with a size lower than 0.125 mm show the most important influence on the workability and rheological properties of concrete (Cepuritis et al., 2016). From the perspective of aggregate types, mortar and concrete prepared with river sand and gravel have relatively low yield stress and plastic viscosity than the mixtures with limestone and crushed aggregates (Hu and Wang, 2007, Lu et al., 2008). As for recycled concrete aggregate (RCA) with high water absorption, experimental results suggest that it is possible to produce SCC with good flowability, excellent passing ability, and stability (Hu et al., 2013). Carro-López et al. (2015) found that replacing natural sand with 20% fine RCA has less influence on the workability and rheological properties of SCC, but further increasing fine RCA replacement significantly reduces the workability of SCC. Besides, the evolution of plastic viscosity over time increases with the replacement of natural sand with fine RCA, as shown in Figure 3.13. However, Güneyisi et al. (2016) suggested that increasing the fine RCA from 0% to 100% increased the flowability of SCC. This might be attributed to the physical properties of the utilized RCA particles. The latter authors also found that the replacement of 50% coarse aggregates by RCA improved the workability of SCC, whereas further increasing the coarse RCA utilization showed less increase in the flowability due to the angular shape of RCA. An empirical relationship between coarse RCA replacements, superplasticizer, w/c, and rheological parameters of the concrete was established by Ait Mohamed Amer et al. (2016) as follows:

w ∆w   X = A (1 + k1RA )  1 + k2 Sp + k3 + k4  (3.31)  c c 

Rheological properties of fresh concrete materials  67

Figure 3.13  Evolution of plastic viscosity over time versus replacement of natural sand by fine RCA. (Data derived from Carro-López et al., 2015.)

Figure 3.14 Relationship between measured rheological properties and calculated parameters by using Eq. (3.31): (a) yield stress and (b) plastic viscosity. (Adapted from Ait Mohamed Amer et al., 2016.)

where X is the yield stress (Pa) or plastic viscosity (Pa.s) of the concrete, A is a parameter depending on the control mixture, R A is the replacement ratio of RCA (%), Sp is the superplasticizer dosage (%), Δw/c is the pre-saturated water-to-cement ratio, and k1–k4 are constants. This simple empirical model provides a useful method to estimate the rheological properties of concrete containing dried or pre-saturated RCA with a high correlation coefficient (Ait Mohamed Amer et al., 2016), as can be observed in Figure 3.14. 3.4 EFFECT OF EXTERNAL FACTORS

3.4.1 Mixing process The mixing process has a significant influence on the dispersion of suspensions, and thus the rheological behavior. In this part, the influences of mixer type, mixing time, and material addition sequence on the rheological properties of mortar and concrete are presented.

68  Rheology of Fresh Cement-Based Materials

Figure 3.15  Effect of mixer type on rheological properties of concrete. (Adapted from Wallevik and Wallevik, 2011.)

Wallevik and Wallevik (2011) stated that the rheological parameters of concrete could be influenced by mixer type. Three different mixers are used, i.e., 50 and 150 l mixers from Maschinenfabrik Gustav Eirich and a typical drum mixer, as shown in Figure 3.15. It can be observed that the drum mixer gave concrete mixtures with higher plastic viscosity than real mixers. With the increase of mixer size, the yield stress of concrete mixtures decreased, but the plastic viscosity showed an opposite behavior. Moreover, they also analyzed the shear rate inside a concrete truck (Wallevik and Wallevik, 2017). They found that the shear rate is dependent on the rotational speed and charge volume of the drum, as well as the rheological parameters of the concrete. Increasing rotational speed increased the shear rate of the drum. If the charge volume is considered alone, the shear rate decreases with the volume exponentially. Struble and Chen (2005) examined the influence of continuous agitation for up to 100 min on the rheological properties of concrete. The yield stress and plastic viscosity increased with time, and after experiencing continuous agitation, the concrete mixture showed lower yield stress and plastic viscosity than the similar mixture without agitation. By theoretical calculations, Li et al. (2004) found that the increase of yield stress and plastic viscosity with time and temperature was more remarkable in stationary state compared to an agitated state. However, overmixing could exert a negative effect on the workability and rheological properties of cement paste and concrete (Dils et al., 2012, Han and Ferron, 2016). In addition, the shear-thinning or thickening intensity of concrete can be altered by subjecting it to continuous agitation. Nehdi and Al Martini (2009) found that concrete mixtures containing superplasticizer exhibited shear-thinning behavior under prolonged mixing and elevated temperature. For the concrete mixtures beyond the saturation dosage of naphthalene sulfonate-based admixture, the rheological behavior shifted from shear-thinning to shear-thickening after subjecting to agitation at a high temperature. Recently, Jiao et al. (2019b) found that fresh mortars with glass fibers tended to show shear-thickening behavior after experiencing a rotational shear mixing. The rheological properties of fresh cementitious materials can be affected by the material addition sequence. Van Der Vurst et al. (2014) evaluated the effects of the addition sequence of materials on rheological properties of SCC. It was found that premixing aggregates with

Rheological properties of fresh concrete materials  69

water increased the plastic viscosity and dynamic yield stress due to the water adsorption effect of aggregates. In contrast, the workability of concretes could be significantly improved by adding aggregates in cement paste. This is in good agreement with the results of França et  al. (2013). Furthermore, França et al. (2016) also examined the influence of polyvinyl alcohol (PVA) fiber and water addition sequences on the mixing and rheological properties of mortar. They found that first mixing water with dry solid parties and then introducing the fibers could save the mixing energy and thus improve the rheological properties of mortar significantly.

3.4.2 Shear history Cementitious suspension is a kind of thixotropic and history-dependent suspension. Lack of steady state during rheological measurements of concrete tends to cause shear-thickening behavior, especially for self-compacting concrete (Feys et al., 2008, Feys et al., 2009). The relaxation period of each rotating velocity influences the implementation of steady state, and thus the measured rheological parameters. When measuring the rheological properties of concrete, non-steady state results in an overestimation of plastic viscosity and underestimation of yield stress (Geiker et al., 2002a). However, a longer rotating time has the potential to lead to the segregation of concrete mixture. To strike a balance between reducing the possible impacts of non-steady state and avoiding segregation, Geiker et al. (2002a) recommended that the rotational speed for unknown concrete should be within 0.05–0.57 rps, and the relaxation time of each rotating speed was 10 s.

3.4.3 Measuring geometry The widely used geometry in cementitious paste includes concentric cylinder, vane, and parallel plate. The measured rheological properties depend on the type and surface of geometry. The potential artifacts, merits, and drawbacks of each geometry have been summarized by Yahia et al. (2017). The measured rheological properties of fresh concrete are also dependent on the rheometer used. The National Institute of Standards and Technology (NIST) carried out a series of experiments to compare the effect of different concrete rheometers such as IBB, BML, and two-point rheometer on the rheological parameters in France (Ferraris and Brower, 2000) and Cleveland (USA) (Ferraris and Brower, 2003). It can be concluded that the absolute yield stress and plastic viscosity are related to the rheometer used, but the rank of rheological parameters is independent of the rheometer. Hočevar et al. (2013) compared the rheological parameters of 26 concrete mixtures measured by ICAR rheometer and ConTec Visco 5. They found that the yield stress and plastic viscosity obtained using the ICAR rheometer was 42% higher and 42% lower than that by ConTec Visco 5, respectively. More details will be provided in Chapter 6.

3.5 SUMMARY Aggregates with larger grain size and rough surface restrict the flow of concrete and increase the yield stress and plastic viscosity compared to cement paste. From a theoretical point of view, the relationship between aggregate volume fraction and plastic viscosity can be predicted by using the Krieger–Dougherty model, while its relationship with yield stress can be estimated by the Chateau–Ovarlez–Trung model. The underlying mechanisms of influence of aggregates on rheological properties can be explained by the excess paste theory.

70  Rheology of Fresh Cement-Based Materials

At relatively low aggregate particle volume fractions, the effect of aggregate size, gradation, and surface texture on the rheological properties of cement-based materials is not significant. At high particle volume fractions, the yield stress and plastic viscosity generally show an exponential increase with the increase of particle volume fraction, which can be explained by the increased collisions and frictions between solid particles at high aggregate volume fractions. For a concrete mixture with a fixed content of cement paste, the higher the aggregate packing density, the lower the fraction of voids, and thus the higher the content of excess paste acting as a lubrication effect. Therefore, the rheological properties of concrete are strongly affected by the packing of aggregate particles. Generally, the particle size of coarse aggregate plays a more significant role in the yield stress compared to the plastic viscosity, and the yield stress and plastic viscosity of concrete mixtures with graded coarse aggregates are significantly lower than that with single-sized coarse aggregates. With regard to particle morphology, aspect ratio and convexity can be used as independent morphological parameters to describe the shape of aggregate particles. Spherical particles generally result in reduced yield stress and plastic viscosity of cement-based materials due to the low uncompacted void content and low friction between particles. The effect of particle shape on the plastic viscosity and consistency index is higher than that on the relative yield stress, and the divergence packing fraction gradually decreases when the particle tends to be more irregular.

REFERENCES Aïssoun, B. M., et al. (2015). “Influence of aggregate characteristics on workability of superworkable concrete.” Materials and Structures, 49(1–2), 597–609. Ait Mohamed Amer, A., et al. (2016). “Rheological and mechanical behavior of concrete made with pre-saturated and dried recycled concrete aggregates.” Construction and Building Materials, 123, 300–308. Barnes, H. A., et al. (1989). An introduction to rheology, Elsevier, Amsterdam, the Netherlands. Bogner, A., et al. (2020). “Early hydration and microstructure formation of Portland cement paste studied by oscillation rheology, isothermal calorimetry, 1H NMR relaxometry, conductance and SAXS.” Cement and Concrete Research, 130, 105977. Bouziani, T., et al. (2012). “Effect of dune sand on the properties of flowing sand-concrete (FSC).” International Journal of Concrete Structures and Materials, 6(1), 59–64. Carro-López, D., et al. (2015). “Study of the rheology of self-compacting concrete with fine recycled concrete aggregates.” Construction and Building Materials, 96, 491–501. Cepuritis, R., et al. (2016). “Crushed sand in concrete – Effect of particle shape in different fractions and filler properties on rheology.” Cement and Concrete Composites, 71, 26–41. Chateau, X., et al. (2008). “Homogenization approach to the behavior of suspensions of noncolloidal particles in yield stress fluids.” Journal of Rheology, 52(2), 489–506. Choi, M. S., et al. (2013). “Prediction on pipe flow of pumped concrete based on shear-induced particle migration.” Cement and Concrete Research, 52, 216–224. Cordeiro, G. C., et al. (2016). “Rheological and mechanical properties of concrete containing crushed granite fine aggregate.” Construction and Building Materials, 111, 766–773. de Larrard, F., and Sedran, T. (2002). “Mixture-proportioning of high-performance concrete.” Cement and Concrete Research, 32, 1699–1704. De Schutter, G., et al. (2018). “Vision of 3D printing with concrete - Technical, economic and environmental potentials.” Cement and Concrete Research, 112, 25–36. Dils, J., et al. (2012). “Influence of mixing procedure and mixer type on fresh and hardened properties of concrete: A review.” Materials and Structures, 45(11), 1673–1683.

Rheological properties of fresh concrete materials  71 Farris, R. J. (1968). “Prediction of the viscosity of multimodal suspensions from unimodal viscosity data.” Transactions of the Society of Rheology, 12(2), 281–301. Ferraris, C. F., and Brower, L. E. (2000). Comparison of Concrete Rheometers: International Test at LCPC (Nantes France) in October, 2000. Ferraris, C. F., and Brower, L. E. (2003). Comparison of Concrete Rheometers: International tests at MB (Cleceland OH, USA) in May, 2003. Feys, D., et al. (2008). “Fresh self compacting concrete, a shear thickening material.” Cement and Concrete Research, 38(7), 920–929. Feys, D., et al. (2009). “Why is fresh self-compacting concrete shear thickening?” Cement and Concrete Research, 39(6), 510–523. França, M. S. D., et al. (2013). “Influence of laboratory mixing procedure on the properties of mortars.” Ambiente Construído, 13(2), 111–124. França, M. S. D., et al. (2016). “Influence of the addition sequence of PVA-fibers and water on mixing and rheological behavior of mortars.” Revista IBRACON de Estruturas e Materiais, 9(2), 226–243. Fuller, W. B., and Thompson, S. E. (1907). “The laws of proportioning concrete.” Transactions of the American Society of Civil Engineers, 59(2), pp. 67–143. Funk, J. E., and Dinger, D. R. (2013). Predictive process control of crowded particulate suspensions: Applied to ceramic manufacturing. Springer Science & Business Media. Geiker, M. R., et al. (2002a). “The effect of measuring procedure on the apparent rheological properties of self-compacting concrete.” Cement and Concrete Research, 32, 1791–1795. Geiker, M. R., et al. (2002b). “On the effect of coarse aggregate fraction and shape on the rheological properties of self-compacting concrete.” Cement, Concrete and Aggregates, 24(1), 3–6. Güneyisi, E., et al. (2016). “Rheological and fresh properties of self-compacting concretes containing coarse and fine recycled concrete aggregates.” Construction and Building Materials, 113, 622–630. Hafid, H., et al. (2016). “Effect of particle morphological parameters on sand grains packing properties and rheology of model mortars.” Cement and Concrete Research, 80, 44–51. Han, D., et al. (2017). “Critical grain size of fine aggregates in the view of the rheology of mortar.” International Journal of Concrete Structures and Materials, 11(4), 627–635. Han, D., and Ferron, R. D. (2016). “Influence of high mixing intensity on rheology, hydration, and microstructure of fresh state cement paste.” Cement and Concrete Research, 84, 95–106. Hočevar, A., et al. (2013). “Rheological parameters of fresh concrete – Comparison of rheometers.” Građevinar, 65(02), 99–109. Hu, C., and de Larrard, F. (1996). “The rheology of fresh high-performance concrete.” Cement and Concrete Research, 26(2), 283–294. Hu, J. (2005). “A study of effects of aggregate on concrete rheology”, Doctoral Thesis, Iowa State University. Hu, J., et al. (2013). “Feasibility study of using fine recycled concrete aggregate in producing selfconsolidation concrete.” Journal of Sustainable Cement-Based Materials, 2(1), 20–34. Hu, J., and Wang, K. (2011). “Effect of coarse aggregate characteristics on concrete rheology.” Construction and Building Materials, 25(3), 1196–1204. Hu, J., and Wang, K. J. (2007). “Effects of size and uncompacted voids of aggregate on mortar flow ability.” Journal of Advanced Concrete Technology, 5(1), 75–85. Iqbal Khan, M., et al. (2016). “Utilization of supplementary cementitious materials in HPC: From rheology to pore structure.” KSCE Journal of Civil Engineering, 21(3), 889–899. Jiao, D., et al. (2017a). “Effects of paste thickness on coated aggregates on rheological properties of concrete.” Journal of the Chinese Ceramic Society, 45(9), 1360–1366. Jiao, D., et al. (2017b). “Effect of constituents on rheological properties of fresh concrete-A review.” Cement and Concrete Composites, 83, 146–159. Jiao, D., et al. (2018). “Mixture design of concrete using simplex centroid design method.” Cement and Concrete Composites, 89, 76–88. Jiao, D., et al. (2019a). “Structural build-up of cementitious paste with nano-Fe3O4 under timevarying magnetic fields.” Cement and Concrete Research, 124, 105857.

72  Rheology of Fresh Cement-Based Materials Jiao, D., et al. (2019b). “Effects of rotational shearing on rheological behavior of fresh mortar with short glass fiber.” Construction and Building Materials, 203, 314–321. Kabagire, K. D., et al. (2017). “Experimental assessment of the effect of particle characteristics on rheological properties of model mortar.” Construction and Building Materials, 151, 615–624. Kabagire, K. D., et al. (2019). “Toward the prediction of rheological properties of self-consolidating concrete as diphasic material.” Construction and Building Materials, 195, 600–612. Krieger, I. M., and Dougherty, T. J. (1959). “A mechanism for non‐Newtonian flow in suspensions of rigid spheres.” Transactions of the Society of Rheology, 3(1), 137–152. Kumar, S., and Santhanam, M. (2003). “Particle packing theories and their application in concrete mixture proportioning: A review.” Indian Concrete Journal, 77(9), 1324–1331. Kwon, S. H., et al. (2013). “Prediction of concrete pumping: Part II—Analytical prediction and experimental verification.” ACI Materials Journal, 110(6), 657–667. Lee, J. H., et al. (2018). “Prediction of the yield stress of concrete considering the thickness of excess paste layer.” Construction and Building Materials, 173, 411–418. Li, T., and Liu, J. (2021). “Effect of aggregate size on the yield stress of mortar.” Construction and Building Materials, 305, 124739. Li, Z., et al. (2004). “Theoretical analysis of time-dependence and thixotropy of fluidity for high fluidity concrete.” Journal of Materials in Civil Engineering, 16(3), 247–256. Lu, G., et al. (2008). “Modeling rheological behavior of highly flowable mortar using concepts of particle and fluid mechanics.” Cement and Concrete Composites, 30(1), 1–12. Mahaut, F., et al. (2008a). “Yield stress and elastic modulus of suspensions of noncolloidal particles in yield stress fluids.” Journal of Rheology, 52(1), 287–313. Mahaut, F., et al. (2008b). “Effect of coarse particle volume fraction on the yield stress and thixotropy of cementitious materials.” Cement and Concrete Research, 38(11), 1276–1285. Mooney, M. (1951). “The viscosity of a concentrated suspension of spherical particles.” Journal of Colloid Science, 6(5), 162–170. Mueller, F. V., et al. (2014). “Linking solid particle packing of Eco-SCC to material performance.” Cement and Concrete Composites, 54, 117–125. Nehdi, M., and Al Martini, S. (2009). “Coupled effects of high temperature, prolonged mixing time, and chemical admixtures on rheology of fresh concrete.” ACI Materials Journal, 106(3), 231–240. Nielsen, L. F. (2001). “Rheology of some fluid extreme composites: Such as fresh self-compacting concrete.” Nordic Concrete Research, 27, 83–94. Noor, M. A., and Uomoto, T. (2004). “Rheology of high flowing mortar and concrete.” Materials and Structures, 37, 512–521. Oh, S., et al. (1999). “Toward mix design for rheology of self-compacting concrete.” Proceedings of Self-Compacting Concrete, Stockholm, 13–14 September 1999, 361–372. Okamura, H., and Ouchi, M. (2003). “Self-compacting concrete.” Journal of Advanced Concrete Technology, 1(1), 5–15. Onoda, G. Y., and Liniger, E. G. (1990). “Random loose packings of uniform spheres and the dilatancy onset.” Physical Review Letters, 64(22), 2727–2730. Ovarlez, G., et al. (2006). “Local determination of the constitutive law of a dense suspension of noncolloidal particles through MRI.” Journal of Rheology, 50(3), 259–292. Park, S., et al. (2018). “Rheological properties of concrete using dune sand.” Construction and Building Materials, 172, 685–695. Quemada, D. (1984). “Models for rheological behavior of concentrated disperse media under shear.” Advances in rheology, 2, 571–582. Reinhardt, H. W., and Wüstholz, T. (2006). “About the influence of the content and composition of the aggregates on the rheological behaviour of self-compacting concrete.” Materials and Structures, 39(7), 683–693. Ren, Q. (2021). “Rheology of cement mortar and concrete as influenced by the geometric features of manufactured sand”, Ghent University. Ren, Q., et al. (2021). “Plastic viscosity of cement mortar with manufactured sand as influenced by geometric features and particle size.” Cement and Concrete Composites, 122, 104163.

Rheological properties of fresh concrete materials  73 Robinson, J. V. (1949). “The viscosity of suspensions of spheres.” The Journal of Physical Chemistry, 53(7), 1042–1056. Roscoe, R. (1952). “The viscosity of suspensions of rigid spheres.” British Journal of Applied Physics, 3, 267. Roussel, N. (2007). “Rheology of fresh concrete: From measurements to predictions of casting processes.” Materials and Structures, 40(10), 1001–1012. Roussel, N., et al. (2010). “Steady state flow of cement suspensions: A micromechanical state of the art.” Cement and Concrete Research, 40(1), 77–84. Roussel, N. (2011). Understanding the rheology of concrete. Elsevier. Roussel, N. (2018). “Rheological requirements for printable concretes.” Cement and Concrete Research, 112, 76–85. Sant, G., et al. (2008). “Rheological properties of cement pastes: A discussion of structure formation and mechanical property development.” Cement and Concrete Research, 38(11), 1286–1296. Santos, A. C. P., et al. (2015). “Experimental study about the effects of granular skeleton distribution on the mechanical properties of self-compacting concrete (SCC).” Construction and Building Materials, 78, 40–49. Struble, L., and Sun, G.-K. (1995). “Viscosity of Portland cement paste as a function of concentration.” Advanced Cement Based Materials, 2(2), 62–69. Struble, L. J., and Chen, C.-T. (2005). “Effect of continuous agitation on concrete rheology.” Journal of ASTM International, 2(9), 1–19. Talbot, A. N., et al. (1923). The strength of concrete: Its relation to the cement aggregates and water. University of Illinois. Toutou, Z., and Roussel, N. (2006). “Multi scale experimental study of concrete rheology: From water scale to gravel scale.” Materials and Structures, 39, 189–199. Van Der Vurst, F., et al. (2014). “Influence of addition sequence of materials on rheological properties of self-compacting concrete.” The 23rd Nordic Concrete Research Symposium, Reykjavik, Iceland, 399–402. Wallevik, J. E., and Wallevik, O. H. (2017). “Analysis of shear rate inside a concrete truck mixer.” Cement and Concrete Research, 95, 9–17. Wallevik, O. H. (2003). “Rheology—A scientific approach to develop self-compacting concrete.” Proceedings of the 3rd International Symposium on Self-Compacting Concrete, Reykjavik, Iceland, 23–31. Wallevik, O. H., and Wallevik, J. E. (2011). “Rheology as a tool in concrete science: The use of rheographs and workability boxes.” Cement and Concrete Research, 41(12), 1279–1288. Wang, X., et al. (2020). “Effect of interparticle action on shear thickening behavior of cementitious composites: Modeling and experimental validation.” Journal of Sustainable Cement-Based Materials, 9(2), 78–93. Westerholm, M., et al. (2008). “Influence of fine aggregate characteristics on the rheological properties of mortars.” Cement and Concrete Composites, 30(4), 274–282. Yahia, A., et al. (2017). “Measuring rheological properties of cement pastes: Most common techniques, procedures and challenges.” RILEM Technical Letters, 2, 129–135. Yammine, J., et al. (2008). “From ordinary rhelogy concrete to self compacting concrete: A transition between frictional and hydrodynamic interactions.” Cement and Concrete Research, 38(7), 890–896. Yuan, Q., et al. (2019). “A feasible method for measuring the buildability of fresh 3D printing mortar.” Construction and Building Materials, 227, 116600.

Chapter 4

Empirical techniques evaluating concrete rheology

4.1 INTRODUCTION Fresh cement-based materials behave as fluids with a yield stress, which is the minimum shear stress for flow to occur (Tattersall and Banfill, 1983, Tattersall, 1991). The fundamental rheological parameters, i.e., yield stress and plastic viscosity, are usually measured using rheological tools such as viscometer and rheometer (Banfill, 2006, Barnes et al., 1989). In situ, the application of rheological apparatus is usually more expensive, difficult, and timeconsuming (Wallevik and Wallevik, 2011). For example, the height of ConTec Viscometer 5, a rheometer that can measure the rheological properties of concrete with slump higher than 120 mm and a maximum aggregate particle size of 22 mm (Wallevik, 2003), is around 2 m, and thus it cannot be used for on-site quality control. For the portable ICAR rheometer, its price is still higher than $20000. In this context, simpler and cheaper conventional workability tests such as the slump test and V-funnel test are more preferred. Although they cannot give the absolute values similar to the rheological parameters, the single-point tests can still be used as effective tools to classify different materials in terms of their ability to be cast (Bouziani and Benmounah, 2013, Gram et al., 2014, Roussel, 2006). For fresh concrete, the yield stress can be correlated to the slump or slump flow, while the plastic viscosity can be predicted by the flow time (Jiao et al., 2017, Lu et al., 2015, Wallevik, 2006). This provides theoretical foundations for evaluating rheological properties by using traditional workability tests. In this chapter, the empirical techniques for evaluating the rheological properties of cement-based materials, such as slump or slump flow, flow time (T50 or V-funnel), and L-box, are presented. The geometry, testing procedure, and data interpretation of the conventional workability tests are illustrated. The mathematical relationships between empirical parameters and rheological properties are also demonstrated. It should be mentioned that the empirical techniques for assessing the rheological properties of cement paste and mortar are beyond the scope of this chapter, with the main emphasis on concrete. 4.2 SLUMP: ASTM ABRAMS CONE The concrete slump test is being used widely by jobsite engineers since a long time ago. Although the construction industry has faced many changes with time, this test is still performed on its old procedure and no changes have been made to this method because of its simple and easily adoptable procedure. It can be easily applied on job site as well as in the laboratory. The basic principle of concrete slump value is a gravity-induced flow of the concrete surface. The slump test is conducted based on the ASTM standard (ASTM-C143/ C143M-12, 2010) or BS EN standard (12350-2, 2009), similar to the Chinese standard (GB/ DOI: 10.1201/9781003265313-4

75

76  Rheology of Fresh Cement-Based Materials

Figure 4.1  ASTM Abrams cone (a) appearance and (b) geometries.

T50080-2002, 2002). This test method is applicable to concrete with a maximum aggregate size of 37.5 mm. The slump of the tested concrete should be within the range of 15–230 mm.

4.2.1 Geometry The apparatus needed for the slump test includes a slump cone, a tamping rod, a scoop, and a measuring device. The appearance of the slump cone and the corresponding geometries are shown in Figure 4.1a and b, respectively. The thickness of the metal cone should be more than 1.5 mm. The base and the top shall be open and parallel to each other and at right angles to the axis of the cone. The diameter of the base and top of the cone is 200 and 100 mm, respectively, and the height is 300 mm. All diameters must be within 3 mm of these diameters. The inside of the cone needs to be smooth and clean, and the cone must be free of dents, adhered mortar, and deformations. For more detailed information about the cone, refer to ASTM-C143/C143M-12 (2010). The tamping rod is a round and straight steel rod with a diameter of 16 mm and a length of approximately 600 mm. A hemispherical tip with a diameter of 16 mm should be attached at the tamping end or both ends. The scoop must be large enough to get a representative sample of concrete and small enough so that it is not spilled during placement into the cone. The measuring device shall have markings in increments of 5 mm or smaller, and the length shall be at least 300 mm.

4.2.2 Testing procedure and parameters The sample of the concrete for the slump test shall be representative of the entire batch, and the selection of the sample is in accordance with ASTM-C172. The testing procedure, as shown in Figure 4.2, shall be as follows:

Empirical techniques evaluating concrete rheology  77

Figure 4.2 (a–f) Main procedures of slump test.

1. Moisten the slump cone to keep concrete from sticking to it. Place the slump cone on a flat, moist, and rigid surface. During concrete filling, the slump cone shall be held firmly by the operator standing on the two-foot pieces. Do not step off until the cone is full and ready to be lifted. 2. Fill the first layer of the concrete up to 70 mm of the cone, 1/3 of the cone by volume. Make sure the concrete is even inside the cone. 3. Rod the layer 25 times throughout its depth by using the tamping rod. Uniformly distribute the strokes over the cross section. Make sure to cover all the surface area inside the cone, slightly angling the rod to get the edges. Do not tap the side of the cone. 4. Fill the second layer to 160 mm, 2/3 of the cone by volume. 5. Rod the layer 25 times, making sure to penetrate the first layer by 25 mm. Again, do not tap the side of the cone. 6. Fill the last layer up to the top, where the concrete is slightly overflowing. Rod the last layer 25 times, and keep the top layer always full. 7. Strike off the excess concrete while keeping the cone steady. Clean around the rim of the cone, making sure it is full. 8. Lift the cone from the concrete carefully in a vertical direction. The lifting process should be within 3–7 s, without lateral or torsional motion. 9. Measure the slump by determining the vertical difference between the top of the mold and the displaced original center of the top surface of the specimen. The slump value is recorded in terms of mm to the nearest 5 mm of subsidence of the specimen during the test. The entire slump procedure needs to be finished within 2.5 min. 10. Clean the equipment thoroughly, and discard the used concrete.

4.2.3 Data interpretation Slumps can be categorized in basic four types with respect to their collapsed shape, i.e., zero slump, true slump, collapsed slump, and shear slump, as shown in Figure 4.3. In zero slump, the fresh concrete does not change its shape during testing. It indicates that a small

78  Rheology of Fresh Cement-Based Materials

Figure 4.3  Types of the slump: (a) zero slump, (b) true slump, (c) collapsed slump, and (d) shear slump.

Table 4.1  Classification of workability based on the concrete slump No. 1 2 3 4

Concrete slump (mm)

Degree of workability

0–25 25–50 50–100 100–175

Very low workability Low workability Medium workability High workability

Table 4.2  Relation between consistency and slump values Slump (mm) Consistency

0–20 Dry

20–40 Stiff

40–120 Plastic

120–200 Wet

200–220 Sloppy

amount of water is used in the mixture design. The true slump is the only type which is allowed under the above-mentioned procedure. The collapse slump where concrete collapses completely indicates that a high amount of water is used to prepare the concrete or the concrete is too wet. Zero slumps and collapse slumps cannot be evaluated by normal tests for slumps because these two do not fall into the workability range, i.e., 15–230 mm. In the case of the shear slump, the top portion of the concrete slump slips away from the side, and ASTM-C143/C143M-12 (2010) advises to repeat the tests with different samples. If the shear slump occurs again, then the concrete mix shall be avoided as it is not sufficiently cohesive. The classification of workability based on concrete slumps is summarized in Table 4.1, and the relation between consistency and slump values is shown in Table 4.2. The slump value can be used to predict the yield stress of fresh concrete, while there is no specific correlation between slump and plastic viscosity (Wallevik, 2006, Zerbino et al., 2008). In the following paragraphs, some representative relationships between slump and yield stress of fresh concrete are presented. By establishing the relationship between yield stress and slump of wet-consistency concrete with a maximum aggregate of 20 mm and slumps from 125 to 260 mm, Murata and Kikukawa (1992) proposed an empirical equation to estimate the yield stress:

τ 0 = 714 − 473log ( SL 10) (4.1)

where τ0 is the yield stress of concrete in Pa, measured by a coaxial cylinder viscometer, SL is the concrete slump in mm, and A and B are experimental constants. From the theoretical

Empirical techniques evaluating concrete rheology  79

Figure 4.4 Representative stress of deformed cone after slump. (Adapted from Schowalter and Christensen, 1998.)

viewpoint, concrete in the cone can be divided into two parts, as shown in Figure 4.4. The shear stress induced by the self-weight of the material is higher than the yield stress of the material, and thus flow occurs in the lower part. However, there is no flow in the upper part. The height of the lower part gradually decreases until the shear stress in this zone reaches the yield stress, where the flow stops. Based on this assumption, Schowalter and Christensen (1998) established a relation between the final total height of the cone and the yield stress, as described by Eq. (4.2), which is independent of the slump cone geometry:



  7 SL = 1 − h0 − 2τ 0 ln   (4.2)  (1 + h0 )3 − 1 

where h0 is the height of the deformed cone. This relation had been successfully validated by Clayton et al. (2003) and Saak et al. (2004) using cylindrical molds. For the ASTM Abrams cone, Hu et al. (1996) established a relationship between the slump and the yield stress by considering the density of the concrete based on a number of numerical simulations, as described by:

SL = 300 − 347

(τ 0 − 212) (4.3) ρ

where SL is the slump in mm, τ0 is the yield stress in Pa, and ρ is the density of the fresh concrete in kg/m3. Besides, through numerical simulating, Roussel (2006) proposed a simple linear relationship between slump and yield stress for traditional concrete with slump values varying from 50 to 250 mm:

SL = 25.5 − 17.6

τ0 (4.4) ρ

where SL is the slump in cm, τ0 is the yield stress in Pa, and ρ is the density of the fresh concrete in kg/m3. There is a good agreement between the BTRHEOM measured yield stress and the theoretical predictions by Eq. (4.4) (Roussel, 2006).

80  Rheology of Fresh Cement-Based Materials

Figure 4.5  Typical relationship between a calculated slump and a measured slump based on Eq. (4.5), where α is 0.0077 mm/(Pa.l), Vmref is 345 l/m3, and τ 0ref is 200 Pa. (Adapted from Wallevik, 2006.)

During the slump flow, the suspended particles must bypass one another. At the stop-flow of the slump, the suspended particles can no longer bypass one another. For a given yield stress, it is found that the stop condition occurs sooner at a relatively low volume fraction of the matrix, due to the reduced distance between particles (Wallevik, 2006). Besides, the relation between slump and yield stress becomes less dependent on the matrix volume fraction when the concrete becomes more workable because of the reduced influence of granular properties by the increased matrix lubrication. After considering the matrix lubrication effect and the granular properties, Wallevik (2006) proposed an empirical relationship between slump and yield stress of concrete, as characterized by:



SL = 300 − 416

(τ 0 + 394 ) + α ρ



0

)(

)

− τ 0ref Vm − Vmref (4.5)

where SL is the slump in mm, τ0 is the yield stress in Pa, ρ is the density of the fresh concrete in kg/m3, α is a constant, τ 0 − τ 0ref is the lubrication effect, and Vm − Vmref is the effect of matric volume fraction. A typical relationship between the calculated slump and the measured slump, with α of 0.0077 mm/(Pa.l), Vmref of 345 l/m3 and τ 0ref of 200 Pa, is shown in Figure 4.5. It can be concluded that Eq. (4.5) provides an excellent prediction of slump for a given yield stress.

(

)

(

)

4.3 SLUMP FLOW AND T50 The slump flow test is generally used to evaluate the workability of high flowable concrete (HFC) and self-compacting concrete (SCC), where the standard slump test is not applicable

Empirical techniques evaluating concrete rheology  81

Figure 4.6  B aseplate and Abrams cone for slump flow test in mm.

anymore, since this kind of concrete cannot maintain its shape when the cone is removed. The slump flow test evaluates the horizontal free flow of the concrete in the absence of obstructions. It is a measure of the filling ability of the concrete, and it can be conducted in the laboratory as well as on-site easily.

4.3.1 Geometry and testing procedure The apparatus for the slump flow test is similar to that of the slump test. The only difference is that a steel base plate marked with circles of 200 and 500 mm at the center is highly recommended, as shown in Figure 4.6. The testing procedure can be found in ASTM-C1611/ C1611M-21 (2009) and EN-BS (2010), according to the following details: 1. Prepare at least 6 L of concrete. 2. Moisten the steel plate and the inside of the slump cone. Put the slump cone on the center of the steel plate, and hold it down firmly. 3. Fill the cone with concrete, without layering, and without any compaction. This process shall be finished within 90 s. 4. Strike off the concrete at the top of the cone with the scraper. Remove the surplus concrete around the base of the cone. 5. Raise the cone vertically within 3 s by a steady upward lift without interruption, and allow the concrete to flow out freely. 6. Measure the time from the beginning of lifting the slump cone to the moment when the maximum diameter reaches 500 mm by using a stopwatch to the nearest 0.1 s, i.e., T50. 7. When the concrete flow stops, measure the largest diameter and the corresponding perpendicular diameter to the nearest 5 mm. The duration from lifting the slump cone to the end of measuring the diameters should be within 40 s. 8. Take the average value of the two diameters as the slump flow to the nearest 5 mm. 9. If the two diameters deviate more than 50 mm, an additional test shall be conducted by using a different sample from the same batch. 10. Clean the steel plate and the slump cone after testing.

82  Rheology of Fresh Cement-Based Materials Table 4.3  SCC classes for different applications (originated from Concrete (2005)) summarized by Zerbino et al. (2008) Class

Range (mm)

Conformity criteria (mm)

Applications

SF1 SF2 SF3

550–650 660–750 760–850

520–700 640–800 740–900

Housing slabs, tunnel linings, piles and deep foundations Waals, columns Very congested structures

4.3.2 Data interpretation The slump flow SF is recorded as the average of the two diameters to the nearest 5 mm, and T50 is the time when the slump flow reaches 500 mm. Note that this flow time can also be designated as T500, and it is lack of precision in the determination due to human eye estimation (Zerbino et al., 2008). The status of the concrete (i.e., segregation or bleeding or not) after flowing should be checked. If most coarse aggregates remain in the center and mortar/ cement pastes gather at the concrete periphery, the concrete can be regarded as severe segregation. For the minor segregation, a border of mortar without coarse aggregate can occur at the edge of the pool of concrete. It should be mentioned that none of the above-mentioned phenomena does not necessarily mean that segregation will not occur due to the timedependent evolution of workability. Based on the range of slump spread flow, SCC can be divided into several classes for different applications, as summarized in Table 4.3. Similar to the slump, the slump spread flow is correlated to the yield stress of the concrete, while the flow time T50 has an apparent relationship with the plastic viscosity (Bouziani and Benmounah, 2013, Cu et al., 2020). Taking the density of concrete into account, Sedran and De Larrard (1999) connected the yield stress to the slump flow from the Abrams cone by using the following empirical equation:

τ 0 = (808 − SF )

ρg (4.6) 11, 740

where SF is the slump flow in mm, g is the gravity acceleration (9.8 m/s2), and ρ is the density of concrete in kg/m3. A typical relationship between the flow time T50 and the plastic viscosity is expressed as follows:

µ=

ρg (0.026SF − 2.39) T50 (4.7) 10,000

where T50 is the flow time in s. It should be mentioned that the constants in Eqs. (4.6) and (4.7) depend on the concrete components or types, and the used rheometer. Similar empirical equations have been obtained by Zerbino et al. (2008), as shown in Figure 4.7, which are independent of the concrete temperature, the mixing energy, the environmental conditions, or the resting time. 4.4 V-FUNNEL TEST FLOW TIME V-funnel test flow time is used to evaluate the filling ability of fresh SCC with a maximum coarse aggregate size of 20 mm through a narrow opening. This test provides a qualitative assessment of the relative viscosity of SCC and indicates the blocking effect caused by

Empirical techniques evaluating concrete rheology  83

Figure 4.7 Typical relationships between rheological properties and slump flow test: (a) yield stress versus slump spread flow and (b) plastic viscosity versus T50. (Adapted from Zerbino et al., 2008.)

Figure 4.8  Geometric parameters of V-funnel.

segregation. A shorter flow time suggests a lower viscosity, whereas a prolonged flow time indicates the blocking susceptibility of the mix.

4.4.1 Geometry The equipment for the V-funnel flow test includes a V-funnel, a stopwatch with an accuracy of 0.1 s, a straightedge, a bucket with a capacity of 12–14 L, and a moist towel. The V-funnel is made of steel, with geometric parameters shown in Figure 4.8. The top of the V-funnel is flat and horizontal, and a momentary releasable and watertight opening gate is installed at the bottom. The V-funnel is placed on vertical support. It should be mentioned that an O-shaped funnel, called O-funnel, can also be used as alternative equipment for the funnel flow time test.

84  Rheology of Fresh Cement-Based Materials

4.4.2 Testing procedure The testing procedure of the V-funnel test is referred to EFNARC (Concrete, 2005), according to the following details:

1. Place the V-funnel on the ground vertically. Make sure the opening top is horizontally positioned. 2. Wet the inner side of the V-funnel using the moist towel, and remove the surplus of water. 3. Close the gate, and place the bucket under the V-funnel to collect the concrete. 4. Fill the V-funnel with a representative SCC completely, without any compaction or vibration. 5. Remove any surplus of the concrete at the top of the funnel using the straightedge. 6. After a resting period of around 10 s, open the gate to let the concrete flow freely. Start to record the stopwatch when the gate opens. 7. Stop the time at the moment when clear space is visible through the opening of the funnel. The reading of the stopwatch is the V-funnel flow time, denoted as T V. Do not move the V-funnel before it is empty. 8. Clean the V-funnel after the test.

4.4.3 Data interpretation The V-funnel flow time T V is the period from opening the gate until the first light enters the gate. It is recorded to the nearest 0.1 s. The V-funnel flow time for SCC should be less than 10 s. To measure the segregation resistance of concrete, the V-funnel is refilled with concrete and allowed to sit for 5 min, and then perform the V-funnel flow test. The segregation resistance is quantified by the flow-through index Sf, as calculated by:

Sf =

T5 − T0 (4.8) T0

where T0 is the initial flow-through time (s) and T5 is the flow-through time after resting for 5 min. There is an acceptable linear relationship between slump flow time T50 and V-funnel flow time T V (Afshoon and Sharifi, 2014, Mohammed et al., 2021, Ouldkhaoua et al., 2019). Compared with T50, the V-funnel flow time is affected by the plastic viscosity, the maximum aggregate size, and the geometric parameter of the bottom opening of the V-funnel (Zerbino et al., 2008). Higher plastic viscosity could possibly be estimated for SCC with coarser aggregates. Nevertheless, the V-funnel flow time can also to a certain degree be correlated to the plastic viscosity of SCC. A representative empirical relationship established by Zerbino et al. (2008) is shown in Figure 4.9, which follows the equation:

µ=

1  T  ln  V  (4.9) 0.013  3.04 

where μ is the plastic viscosity (Pa.s) and T V is the V-funnel flow time (s). Besides, a linear relationship between plastic viscosity and V-funnel flow time had also been established (Boukhelkhal et al., 2016, Ouldkhaoua et al., 2019).

Empirical techniques evaluating concrete rheology  85

Figure 4.9 Typical relationship between plastic viscosity and V-funnel flow time. (Adapted from Zerbino et al., 2008.)

Figure 4.10 Geometric parameters of standard L-box.

4.5 OTHER METHODS In addition to the above-mentioned tests, there are some other available empirical workability tests to evaluate the rheological properties of fresh concrete. The apparatus, testing procedure, and data interpretation of some typical empirical methods, including L-box, LCPC box, V-funnel coupled with a horizontal channel, and J-ring, are illustrated.

4.5.1 L-box L-box is the reference method for evaluating the passing ability of SCC and its passing profile along with reinforcement. The geometric parameters of the L-box are shown in Figure 4.10. Two types of gates, one with 3 smooth bars and the other with 2 smooth bars, can be used. The gaps between neighboring bars are 41 and 59 mm, respectively.

86  Rheology of Fresh Cement-Based Materials

The testing procedure of the L-box test is as follows (Nguyen et al., 2006, Skarendahl et al., 2000):

1. Wet the internal surface of the L-box, and then place it in a stable and level position. 2. Fill the vertical part of the L-box with 12.7 L representative fresh SCC. 3. Wait for 1 min to check whether the concrete is stable or not. 4. Lift the gate slowly, and let the concrete flow into the horizontal part freely. 5. When the concrete stops moving, measure the average distance of the top of the concrete in the vertical part (h1) as well as the one at the end of the box (h2). Three positions with one at the center and two at each side shall be recorded. 6. The L-box value is calculated as h2 /h1. If the L-box value is equal to 1, the concrete is perfectly flowable. Conversely, if the concrete is too stiff, then the L-box value is equal to 0. For SCC, the L-box value varies from one country to another in the range of 0.60–1. The L-box value can be linked to the rheological properties of SCC. Given that the fluid is homogeneous and the gate is lifted slowly, the relationship between the L-box measured parameters and the yield stress/density ratio of the material can be expressed as follows (Nguyen et al., 2006):

h1 − h2 =

τ 0 L0  l0 L0 2  τ +  + B 0 (4.10) l0  ρ g  V ρg

where L 0 and l 0 are the length and width of the L-box channel, respectively; V is the total volume of the sample; τ0 is the yield stress; g is the gravity acceleration; ρ is the density of concrete; and B is a dimensionless constant depending on the geometry of the L-box. Note that Eq. (4.10) works well for limestone powder suspensions, but it cannot be validated for concrete due to the lack of absolute value of the yield stress. Nevertheless, this correlation has a potential practical application in predicting the yield stress of SCC. For example, a plot of L-box value and yield stress/density ratio, as shown in Figure 4.11, is a practical tool for evaluating the yield stress of SCC in on-site works (Chamani et al., 2014). Alternatively, the relationship can be expressed by:

Figure 4.11 Relationship between L-box value and yield stress/density ratio. (Adapted from Chamani et al., 2014.)

Empirical techniques evaluating concrete rheology  87

Figure 4.12  LCPC box for measuring the yield stress of SCC.



3 3 3  h2  h2   h2   h2  ≤ 0.5 91.3   − 175.3   − 39.6   + 104.8, 0 ≤  h1   h1   h1  h1  τ0 = 340.5  (4.11) 3 3 3 SG  h2  h2   h2   h2  ≤1 340.5   − 609.8   + 214.1   + 55.2, 0.5 <       h h h h1 1 1 1 

where SG is the specific gravity of the sample.

4.5.2 LCPC box Roussel (2007) proposed a cheap and simple technique, the LCPC box, to measure the yield stress of SCC. The LCPC box test for SCC is shown in Figure 4.12. The length and the width of the LCPC box are 1.2 and 0.2 m, respectively. The height of the LCPC box is 0.15 m. The testing procedure of the LCPC box test is as follows: 1. Fill a pre-wetted bucket with 6 L tested SCC. 2. Slowly pour the SCC at the end of the LCPC box (see Figure 4.12). The pouring process shall be within 30 s. 3. Wait until the concrete flow stops. 4. Measure the spread length L, the initial height Hi, and the final height Hf of the concrete in the LCPC box. 5. Calculate the yield stress based on the following equation:



τ0 =

ρ gl0  l 0  l 0 + 2H f ( H i − H f ) + ln  2L  2  l 0 + 2H i

   (4.12) 

where ρ is the density of the concrete, l 0 is the width of the L-box, L is the length of the flow, g is the acceleration gravity, Hi and Hf are the initial and final height, respectively. A representative correlation between the spread length and the ratio of yield stress/specific gravity is shown in Figure 4.13. The LCPC box test is a cheap, simple, and precise measurement of the yield stress of any SCC. However, it is highly dependent on the operator, and the use of a bucket can possibly influence the concrete flow (Benaicha et al., 2015).

88  Rheology of Fresh Cement-Based Materials

Figure 4.13  Representative correlation between the spread length from LCPC box and the ratio of yield stress/specific gravity. (Adapted from Roussel, 2007.)

Figure 4.14  V-funnel coupled with a horizontal channel. (Adapted from Benaicha et al., 2015.)

4.5.3 V-funnel coupled with a horizontal channel Based on the LCPC box test, Benaicha (2013) and Benaicha et al. (2015) modified the traditional V-funnel coupled with a horizontal channel to predict the plastic viscosity of concrete. The modified device is shown in Figure 4.14. The length of the channel is 0.90 m, the same length as the slump flow table. The channel has the same width as L-box, i.e., 0.20 m. The height of the channel is 0.16 m. The flow time and the flow profile of concrete in the channel can be determined at any time, which can be used to calculate the plastic viscosity of concrete, as described by:



2   dt  8Hτ 0  2(z tan α + d) + e   dt  z tan α + d  − + + − 1 ρ gz ρ − ξ          3  (z tan α + d) ⋅ e  d  dz   2dz µp = (4.13) 4 16H  [ 2(z tan α + d) + e ]    π  [(z tan α + d) ⋅ e ]3 

Empirical techniques evaluating concrete rheology  89

Figure 4.15 Definition of geometric parameters of the V-funnel. (Adapted from Benaicha et al., 2015.)

where d, H, e, and α are the geometric parameters of the V-funnel, as shown in Figure 4.15. dz/dt is the flow velocity of concrete, μp is the plastic viscosity, τ0 is the yield stress, ρ is the concrete density, and g is the gravity. By using MATLAB® program with the Runge-Kutta method, the plastic viscosity of concrete can be obtained from the V-funnel flow time and length. A simple expression of Eq. (4.13) was described by Benaicha et al. (2017):

3



µp =

−5

1  Q   L −4/5  t  (4.14) ρ − ρair ) g    (  e   0.95  3

where ρ and ρair are the density of the concrete and air, respectively; Q is the volumetric flow rate; e is the channel width; L is the length of the flow; and t is the flow time. Experimental results with more than 100 different compositions showed that the calculated plastic viscosity had an excellent correlation with the measured ones using a rheometer, with a coefficient of determination of 0.9224 (Benaicha et al., 2015). The V-funnel coupled with a horizontal channel is an efficient, simple, and economical tool to characterize the plastic viscosity of concrete.

4.5.4 J-ring J-ring test is an alternative method for evaluating the filling ability and passing ability of high flowability concrete, especially SCC. The apparatus of the J-ring test includes all the apparatuses for slump flow test and a J-ring with a rectangular section of 30 mm × 50 mm and a diameter of 300 mm open steel ring drilled vertically with holes to accept threaded sections of reinforcing bars generally at a spacing of 48 ± 2 mm, as shown in Figure 4.16. The testing procedure of the J-ring test is referred to De Schutter (2005), according to the following steps: 1. Moisten the inside of the slump cone and base plate. 2. Place the J-ring centrally on the base plate and the slump cone centrally inside the J-ring.

90  Rheology of Fresh Cement-Based Materials

Figure 4.16 J -ring test apparatus. (Adapted from De Schutter, 2005.)

3. Fill the slump cone with about 6 L concrete by using a scoop without any external compacting action. Simply strike off the concrete level with a trowel, and remove all surplus concrete. 4. After a short rest (no more than 30 s for cleaning), raise the cone vertically and allow the concrete to flow out through the J-ring. 5. Record the time when the front of the concrete first touches the circle of a diameter of 500 mm, i.e., T50J. The test is completed when the concrete flow is ceased. 6. Measure the largest diameter (dmax) and the one perpendicular to it (dperp), to the nearest 5 mm. Calculate the average diameter. 7. Measure the difference in height between the concrete surface at the central position (Δh0) and the four positions outside the J-ring, Δhx1, Δhx2 , Δhy1, and Δhy2 , as shown in Figure 4.16. Calculate the average of the difference in height at four locations in mm. 8. Clean the base plate, the J-ring, and the cone after testing. Three parameters can be obtained from the J-ring flow test, including J-ring flow spread (S J), J-ring flow time (T50J), and J-ring blocking step (BJ), which can be expressed by:

SJ =

dmax + dperp (4.15) 2



BJ =

∆hx1 + ∆hx2 + ∆hy1 + ∆hy 2 − ∆h0 (4.16) 4

S J is expressed in mm to the nearest 5 mm, indicating the restricted deformability of SCC due to the blocking effect. T50J is recorded in s to the nearest 0.1 s, which indicates the rate of deformation. BJ is expressed in mm to the nearest 1 mm, representative of the blocking effect. The acceptable difference in height between inside and outside should be between 0 and 10 mm. Although no quantitative equation is established between the J-ring measured parameters and rheological properties, the J-ring flow time and J-ring spread flow can be empirically

Empirical techniques evaluating concrete rheology  91

Figure 4.17 J -ring flow time T50J versus plastic viscosity of SCC. (Adapted from Abo Dhaheer et al., 2015.)

correlated to the plastic viscosity and yield stress, respectively. A typical relationship between J-ring flow time T50J and plastic viscosity of SCC is shown in Figure 4.17. It can be seen that a quadratic relationship between T50J and plastic viscosity is observed. In the case of J-ring spread flow, it can be correlated to the spreading flow measured from the slump flow test, as shown in Figure 4.17. More details can be found in the works of Barroqueiro et al. (2019), Daoud and Kabashi (2015), and Pradoto et al. (2016). 4.6 SUMMARY The fundamental rheological parameters including yield stress and plastic viscosity are usually measured using rheological tools. In situ, however, the application of rheological apparatus is usually more expensive, difficult, and time-consuming. Therefore, simpler and cheaper conventional workability tests such as the slump test and V-funnel tests can be used as effective tools to classify different materials in terms of their ability to be cast. The geometry, testing procedure, and data interpretation of the conventional workability tests (including slump, slump spread flow and flow time, V-funnel, L-box, and J-ring) are illustrated. The mathematical relationships between empirical parameters and rheological properties are also demonstrated. For fresh concrete, the yield stress can be correlated to the slump or slump flow, while the plastic viscosity can be predicted by the flow time (T50, V-funnel flow time, or J-ring flow time). REFERENCES Abo Dhaheer, M. S., et al. (2015). “Proportioning of self-compacting concrete mixes based on target plastic viscosity and compressive strength: Part II - experimental validation.” Journal of Sustainable Cement-Based Materials, 5(4), 217–232. Afshoon, I., and Sharifi, Y. (2014). “Ground copper slag as a supplementary cementing material and its influence on the fresh properties of self-consolidating concrete.” The IES Journal Part A: Civil & Structural Engineering, 7(4), 229–242.

92  Rheology of Fresh Cement-Based Materials ASTM-C143/C143M-12 (2010). “Standard test method for slump of hydraulic-cement concrete.” Annual Book of ASTM Standards, American Society for Testing and Materials, Philadelphia, USA. ASTM-C172. “Practice for sampling freshly mixed concrete.” Annual Book of ASTM Standards, American Society for Testing and Materials, Philadelphia, USA. ASTM-C1611/C1611M-21 (2009). “Standard test method for slump flow of self-consolidating concrete”, Annual Book of ASTM Standards, American Society for Testing and Materials, Philadelphia, USA. Banfill, P. (2006). “The rheology of fresh cement and concrete-rheology review.” British Society of Rheology, 61, 130. Barnes, H. A., et al. (1989). An introduction to rheology. Elsevier, Amsterdam, the Netherlands. Barroqueiro, T., et al. (2019). “Fresh-state and mechanical properties of high-performance self-compacting concrete with recycled aggregates from the precast industry.” Materials (Basel), 12(21), 3565. Benaicha, M. (2013). Rheological and mechanical characterization of concrete: New approach. LAP Lambert Academic Publishing, Chisinau, Moldova. Benaicha, M., et al. (2015). “New approach to determine the plastic viscosity of self-compacting concrete.” Frontiers of Structural and Civil Engineering, 10(2), 198–208. Benaicha, M., et al. (2017). “Theoretical calculation of self-compacting concrete plastic viscosity.” Structural Concrete, 18(5), 710–719. Boukhelkhal, A., et al. (2016). “Effects of marble powder as a partial replacement of cement on some engineering properties of self-compacting concrete.” Journal of Adhesion Science and Technology, 30(22), 2405–2419. Bouziani, T., and Benmounah, A. (2013). “Correlation between v-funnel and mini-slump test results with viscosity.” KSCE Journal of Civil Engineering, 17(1), 173–178. BS EN standard 12350-2. (2009). “Testing fresh concrete: Slump-test.” Part 2: British Standards. Chamani, M. R., et al. (2014). “Evaluation of SCC yield stress from L-box test using the dam break model.” Magazine of Concrete Research, 66(4), 175–185. Clayton, S., et al. (2003). “Analysis of the slump test for on-site yield stress measurement of mineral suspensions.” International Journal of Mineral Processing, 70(1–4), 3–21. Cu, Y. T. H., et al. (2020). “Relationship between workability and rheological parameters of selfcompacting concrete used for vertical pump up to supertall buildings.” Journal of Building Engineering, 32, 101786. Daoud, O., and Kabashi, T. (2015). “Production and fresh properties of powder type self—­compacting concrete in Sudan”, in Concrete repair, rehabilitation and retrofitting IV., Edited by Dehn, Beushausen, Alexander and Moyo. CRC Press, Leipzig, Germany, pp. 99–99. De Schutter, G. (2005). “Guidelines for testing fresh self-compacting concrete.” European Research Project. Concrete, S.C., (2005). “The European guidelines for self-compacting concrete.” BIBM, et al., pp. 50–52. EN-BS (2010). 12350-8 (2010). Testing fresh concrete self-compacting concrete. Slump-flow test. British Standards Institute, London, United Kingdom. Concrete, S.C., (2005). “The European guidelines for self-compacting concrete.” BIBM, et al., pp. 47–59. GB/T50080-2002 (2002). “Standard for test method of performance on ordinary fresh concrete.” Chinese Standard, China Academy of Building Research, Guangdong, China. Gram, A., et al. (2014). “Obtaining rheological parameters from flow test — Analytical, computational and lab test approach.” Cement and Concrete Research, 63, 29–34. Hu, C., et al. (1996). “Validation of BTRHEOM, the new rheometer for soft-to-fluid concrete.” Materials and Structures, 29(10), 620–631. Jiao, D., et al. (2017). “Effect of constituents on rheological properties of fresh concrete-A review.” Cement and Concrete Composites, 83, 146–159. Lu, C., et al. (2015). “Relationship between slump flow and rheological properties of self compacting concrete with silica fume and its permeability.” Construction and Building Materials, 75, 157–162.

Empirical techniques evaluating concrete rheology  93 Mohammed, A. M., et al. (2021). “Experimental and statistical evaluation of rheological properties of self-compacting concrete containing fly ash and ground granulated blast furnace slag.” Journal of King Saud University - Engineering Sciences. Murata, J., and Kikukawa, H. (1992). “Viscosity equation for fresh concrete.” ACI Materials Journal, 89(3), 230–237. Nguyen, T. L. H., et al. (2006). “Correlation between L-box test and rheological parameters of a homogeneous yield stress fluid.” Cement and Concrete Research, 36(10), 1789–1796. Ouldkhaoua, Y., et al. (2019). “Rheological properties of blended metakaolin self-compacting concrete containing recycled CRT funnel glass aggregate.” Epitoanyag-Journal of Silicate Based & Composite Materials, 71(5):154–161. Pradoto, R., et al. (2016). “Fly ash-nano SiO2 blends for effective application in self-consolidating concrete.” Proceedings of 8th International RILEM Symposium on Self-Compacting Concrete, Washington DC, 299–308. Rezania, M., et al. (2019). “Experimental study of the simultaneous effect of nano-silica and nanocarbon black on permeability and mechanical properties of the concrete.” Theoretical and Applied Fracture Mechanics, 104: 102391. Roussel, N. (2006). “Correlation between yield stress and slump: Comparison between numerical simulations and concrete rheometers results.” Materials and Structures, 39(4), 501–509. Roussel, N. (2007). “The LCPC BOX: A cheap and simple technique for yield stress measurements of SCC.” Materials and Structures, 40(9), 889–896. Saak, A. W., et al. (2004). “A generalized approach for the determination of yield stress by slump and slump flow.” Cement and Concrete Research, 34(3), 363–371. Schowalter, W. R., and Christensen, G. (1998). “Toward a rationalization of the slump test for fresh concrete: Comparisons of calculations and experiments.” Journal of Rheology, 42(4), 865–870. Sedran, T., and De Larrard, F. (1999). “Optimization of self compacting concrete thanks to packing model.” RILEM Proceedings, 7, 321–332. Skarendahl, A., et al. (2000). “Self-compacting concrete-state-of-the-art report of RILEM TC 174SCC.” RILEM report, 23. Suliman, M. O., et al. (2017). “Effects of stone cutting powder (Al-Khamkha) on the properties of self-compacting concrete.” World Journal of Engineering and Technology, 05(04), 613–625. Tattersall, G. H. (1991). Workability and quality control of concrete. CRC Press, Boca Raton, FL. Tattersall, G. H., and Banfill, P. F. (1983). The rheology of fresh concrete. Pitman Books Limited, London, England. Wallevik, J. E. (2006). “Relationship between the Bingham parameters and slump.” Cement and Concrete Research, 36(7), 1214–1221. Wallevik, O. H. (2003). “Rheology—A scientific approach to develop self-compacting concrete.” Proceedings of the 3rd International Symposium on Self-Compacting Concrete, Reykjavik, 23–31. Wallevik, O. H., and Wallevik, J. E. (2011). “Rheology as a tool in concrete science: The use of rheographs and workability boxes.” Cement and Concrete Research, 41(12), 1279–1288. Zerbino, R., et al. (2008). “Workability tests and rheological parameters in self-compacting concrete.” Materials and Structures, 42(7), 947–960.

Chapter 5

Paste rheometers

5.1 INTRODUCTION TO THE RHEOLOGY OF CEMENT PASTE Rheology is an important branch of natural science. It is widely used in suspension, polymer, food, coating, and cosmetic industries to evaluate the rheological properties of fluids. Principles of rheology applied to cement-based materials will provide us with an important means to accurately characterize the rheological properties of concrete. The good rheology of concrete is beneficial to pouring, forming, and transmission under complex conditions and even to the development of later strength and durability, especially in the field of pumped and sprayed concrete (Morinaga, 1973, Glab et al., 2020). At the same time, for 3D-printed concrete, to ensure its extrudability, uniformity, and buildability, the rheological properties of the concrete need to be accurately controlled within an appropriate range (Mca et al., 2020). In addition, rheological measurements have become a means of monitoring changes in the microstructure of cement-based materials (Wallevik, 2009). Cement paste is the most important component of concrete. Therefore, the rheological behavior of paste is the basis of concrete rheological properties (Ferraris and Gaidis, 1992). Fresh cement paste is a heterogeneous material that can be regarded as a suspension of cement particles in water. The particle size ranges from nanometers (e.g., hydration products and polymers) to micrometers (e.g., cement and minerals). The mechanical interaction between solids and the chemical hydration between cementitious particles and water together affect their rheological properties (Liu et al., 2017). For a long time, most cement pastes have been considered as Bingham fluids, exhibiting yield stress (Cepuritis et al., 2019). Because of the friction between cement particles, it also has viscosity. The Bingham model describes the linear relationship between shear stress and shear rate through two rheological parameters. However, the modification of chemical admixtures and mineral additives makes the rheological behavior of cement paste more complicated (Ferraris et al., 2001, Alonso et al., 2007, Hanehara and Yamada, 1999, Khayat, 1998). The stress–strain relationship of the slurry will appear in a nonlinear form, including shear thinning and shear thickening (Ma et al., 2016, Deng et al., 2013, Maybury et al., 2017, Yang et al., 2018). Shear thinning and shear thickening may occur simultaneously, depending on which one is dominant (Liu et al., 2017). In particular, cement paste exhibits thixotropy, which usually shows that under the action of shear stress, the viscosity decreases with time, and it recovers after the removal of the shear stress. The thixotropy of cement paste increases over time due to the process of hydration (Roussel et al., 2012, Erdem et al., 2015). Therefore, thixotropy may cause an overestimation of rheological parameters. In addition, the measurement results of the rheometer are also affected by other factors, including system friction, thermal expansion, temperature, humidity (Schüller and Salas-Bringas, 2007, Nehdi and Rahman, 2004, Lewis et al., 2000), and pressure (Hyun et al., 2002, Proske et al., 2020). Therefore, under DOI: 10.1201/9781003265313-5

95

96  Rheology of Fresh Cement-Based Materials

the most suitable rheological model, obtaining accurate rheological parameters of cement slurry and controlling and adjusting the number of mineral admixtures used in concrete will have good application prospects (Chen and Rothstein, 2004, Sun et al., 2007).

5.2 RHEOMETERS FOR CEMENT PASTE

5.2.1 Narrow gap coaxial cylinder rheometer 5.2.1.1 Geometry The coaxial cylinder rheometer usually includes an inner cylinder and an outer cylinder. During the measurement process of torque, the inner cylinder or the outer cylinder maintains rotation, while the other cylinder remains stationary. For cement paste, ASTM C 1749 (ASTM C1749-17a) gives procedures for measuring rheological parameters under narrow and wide gap conditions, and it points out that for a narrow gap, the ratio between inner and outer diameters is greater than or equal to 0.92. The gap should be 10 times greater than the diameter of the cementitious particles, approximately 0.4 mm. Generally, the gap of the rheometer is slightly larger than this value by approximately 1 mm. To calculate the shear stress and shear rate from the known torque and speed, we assume that the flow of the slurry between the inner and outer cylinders is simple. First, there is no slippage between the surface of the cylinder and the slurry. Second, the flow of the slurry along the radius is laminar, no flow occurs in the vertical direction, and there is only a velocity component along the angle. This is the same assumption as the concrete rheometer. Under narrow gap conditions, the most important thing is that the shear rate can be considered constant. Therefore, the rheological parameters can be directly obtained from the shear stress and shear rate at the rotating drum. This assumption is not applicable under wide gaps because of the more complicated flow behavior (Liu et al., 2020). To reduce the influence of wall slippage, the traditional cylindrical rotor can be replaced by a vane rotor. The gap between the blades will be filled by the slurry, which is theoretically the same as a cylindrical rotor. The effectiveness of the blade rotor instead of the cylindrical rotor is verified by numerical simulation (Zhu et al., 2010). There are two main advantages of using blades, which are determined by their geometry. First, when the four-blade rotor rotates, the outer end is not a completely smooth round surface, so there is no wall slip. Second, due to the influence of its volume and geometry, the structure of the sample is much less damaged by inserting the blade (Alderman et al., 1991). However, some researchers have found that the position where the shear surface occurs when measuring the Bingham fluid using a four-blade rotor is slightly larger than the radius of the blade rotor (Keentok et al., 1985). 5.2.1.2 Measurement principle For a narrow gap, the shear rate at the inner cylinder can be obtained from the gap and angular velocity between the inner and outer cylinders, which can be expressed as Eq. (5.1):

γ =

ω R1 (5.1) R2 − R1

where γ is the shear rate, ω is the angular velocity of rotation, R1 is the radius of the inner cylinder, and R 2 is the radius of the outer cylinder.

Paste rheometers  97

The shear stress at the inner cylinder can be calculated by Eq. (5.2):

τ=

T (5.2) 2πR12 h

where τ is the shear stress at the inner cylinder, T is the torque value, and h is the rotor height. Therefore, the rheological parameters can be obtained by the intercept and slope of the linear relationship between the shear stress at the inner cylinder and the shear rate. The unit of yield stress is Pa, and the unit of plastic viscosity is Pa.s. For wide gap conditions, rheological parameters can be obtained by the Reiner–Riwlin equation. The shear rate is calculated from Eq. (5.3):

γ = r

dω (5.3) dr

The shear stress at radius r is given by Eq. (5.4):

τ=

T (5.4) 2πr 2 h

For the Bingham fluids, the linear relationship between shear stress and shear rate can be expressed as Eq. (5.5):

τ = τ 0 + µγ (5.5)

Substitute Eqs. (5.4) and (5.5) into Eq. (5.3) as Eq. (5.6):

τ dω T = − 0 (5.6) 2 dr 2πr hµ µ

r

Integrating Eq. (5.7) is shown in Eq. (5.8):





0



dω =

Ω=



R2

R1

 T τ0   2πr 3 hµ − µ r  dr (5.7)

T  1 1  τ 0  R2  − − ln (5.8) 4πhµ  R12 R22  µ  R1 

However, at a certain speed of the inner cylinder, the slurry has an unsheared area, that is, a dead zone. At this time, the radius of the sheared area is smaller than the radius of the outer cylinder, and it is often necessary to correct the torque or speed during the calculation. Therefore, the abovementioned Reiner–Riwlin equation can be rewritten as Eq. (5.9):

Ω=

1  τ T  1  Rplug  − 2  − 0 ln  (5.9)  2 4πhµ  R1 Rplug  µ  R1 

98  Rheology of Fresh Cement-Based Materials

The radius of the plug flow area Rplug can be expressed as Eq. (5.10):

Rplug =

T (5.10) 2πhτ 0

Koehler and Fowler (2004) compared three methods of dead zone correction with existing concrete rheometers, namely Point Elimination Method, Independent Yield Stress Method, and Effective Annulus Method. These methods can also be used in slurry rheometers.

5.2.1.3 Measuring errors and artifacts In the measurement of a coaxial cylindrical rheometer, the wall slip effect of a smooth cylindrical rotor has been widely accepted. This is due to a thin layer with a lower particle concentration produced near the wall. Accordingly, the actual shear rate of the slurry decreases, and the test produces a large error (Harboe et al., 2012). The most common way to reduce wall slip is to increase the roughness of the rotor surface. The thickness of the thin slip layer is also affected by the physical properties of the test fluid. For example, the particles of suspension in contact with the wall of the rheometer are affected by their radius (Haimoni and Hannant, 1988). Figure 5.1a shows the remarkable feature that the stress changes in a zigzag shape when wall slip occurs (Saak et al., 2001). Figure 5.2 shows some surface-treated rotors and the inner wall of the outer cylinder, including surface protrusions or grooves and blade rotors, sandblasting, and jagged surfaces. The wall slip velocity is considered to be the difference between the wall velocity and the fluid velocity near the wall. When using a roughened rotor or even a bladed rotor instead of a cylindrical rotor, the results of rheological measurements may be different due to the wall slip effect (Nehdi and Rahman, 2004, Haimoni and Hannant, 1988). In terms of experiments, the coaxial cylindrical rheometer uses the roughened-surface rotor to obtain higher torque or shear stress than the cylindrical rotor (Saak et al., 2001, Harboe et al., 2011, Haimoni and Hannant, 1988). Figure 5.1b (Saak et al., 2001) shows the stress– speed relationship between the blade rotor and the cylindrical rotor. Figure 5.3 shows representative shear stress–shear rate curves under smooth, slotted, and bladed rotors. In addition, researchers have used numerical simulation methods to prove that the roughening of the surface makes the impact of wall slip no longer significant, and higher accuracy is obtained (Wang et al., 2010). Zhu et al. (2010) compared the rheological parameters obtained at different shear rates with the experimental results and reported that the blade rotor can approximately replace the cylindrical rotor to measure the rheology of the Bingham fluid. The wall slip of the cylindrical rotor underestimates the viscosity when slip occurs. This is the same as the simulation result of Barnes (1990). However, the blade may not be completely equivalent to a cylinder, and the secondary flow between the blade gaps affects the viscosity measurement. In the measurement of wall slip velocity, OrtegaAvila et al. (2016) used particle image velocimetry to measure the velocity of gel flowing in the annular gap between two concentric cylinders. The theoretical calculation of wall slip velocity has a certain mathematical foundation, and its correctness has been verified through experiments (Yoshimura and Prud’homme, 1988). Therefore, in most cases, surface roughening is effective in reducing the effect of wall slip, but how to define the most appropriate roughness conditions remains to be studied. ASTM (ASTM C 1749-17a) provides a standardized method for measuring cement slurries under rough surface conditions.

Paste rheometers  99

Figure 5.1 Representative stress growth curves (a) and the influence of rotational speed (b) for concentric cylinders and vanes. The rotational speed of the stress growth test is 0.01 rad/s, and the w/c ratio is 0.3 (Saak et al., 2001).

5.2.2 Plate–plate rheometer 5.2.2.1 Geometry The parallel-plate rheometer usually includes a rotating upper plate and a stationary lower plate, and the gap between the two parallel plates can be adjusted as needed. It will be filled with slurry, and the torque applied to the slurry by the upper plate will be measured. When measuring cement paste, the gap between the two plates is not always guaranteed to be a small gap, so that the shear rate is not constant. The largest at the upper board, and the smallest at the lower board. Based on the assumption of laminar flow, the shear rate varies linearly between the upper and lower plates. In addition, whether in a smooth or nonsmooth plate, the speed of the upper plate will also cause a change in rheological parameters when the upper plate is lowered because the speed of the squeezing fluid will cause different degrees of water migration (Cardoso et al., 2015). The upper plate can be replaced with a conical plate. The greatest advantage compared to a flat plate is that the shear rate at different heights is constant, and the cone plate reduces the volume of the sample. However,

100  Rheology of Fresh Cement-Based Materials

Figure 5.2  Concentric cylinder test geometry: (a) from left: smooth rotor, grooved rotor, 4-blade vane and (b) from left: grooved cup, smooth cup (Shamu and Hkansson, 2019).

Figure 5.3 Comparison of flow curves for cement grout with w/c of 0.6 (a) and 0.8 (b) obtained in grooved, smooth, and vane geometries (Shamu and Hkansson, 2019).

the centrifugal effect at the conical plate makes the slurry unable to adhere well to the wall. Besides, it is more sensitive to the particle size, usually greater than ten times the maximum particle size. Compared with the coaxial cylinder rheometer, the parallel-plate rheometer is more suitable for high-viscosity slurries. On the one hand, it is easier to add high-viscosity samples between the parallel plates and will not be affected by geometry. On the other hand, the parallel-plate measurement requires fewer samples, and the maximum shear rate can be changed by changing the diameter or gap. However, in the coaxial cylinder rheometer, the low-viscosity fluid can fill the gap between the inner and outer cylinders well and at the same time can better control the temperature of the sample during the measurement. There is a

Paste rheometers  101

Figure 5.4 (a–c) Parallel-plate test geometry (Pawelczyk et al., 2020).

temperature difference between the samples in the gap between the parallel plates, and it is often positively correlated with the gap. Cone-plate conditions require the least amount of sample. ASTM (C 1749-17a) provides a test procedure for parallel plates. To reduce the influence of wall slippage, methods such as sandpaper bonding, sandblasting, serrating, and grooving are also used. Figure 5.4 shows some parallel-plate test geometries. The position of the parallel plate should be first determined at the beginning of the test. The size of the gap has a significant impact on the rheological parameters (Carotenuto and Minale, 2013, Nickerson and Kornfield, 2005). It is determined by the maximum particle size of the particles in the tested slurry. When the gap is narrowed, the friction between the particles and the plate will increase, resulting in a higher normal force and the changing of the flow state. If the gap is too large, it cannot be effectively filled by slurry (Ferraris et al., 2014). To ensure that the initial measurement is at the zero position, it is more difficult to determine under a rough surface because of its geometric height limitation, which will have a greater impact on the calculation of rheological parameters. 5.2.2.2 Measurement principle The torque and speed data are obtained by measurement, and the linear relationship between torque and speed is obtained by fitting. Take the most common Bingham model as an example, the relationship between torque and shear stress is as follows:

T=



τ dA r =



R

0

2πr 2 (τ 0 + µγ ) dr (5.11)

We can obtain the shear rate at the edge of the upper plate by Eq. (5.12):

γ =

ωR (5.12) h

102  Rheology of Fresh Cement-Based Materials

Afterward, we can convert the variable in the integral into a shear rate:

T=

2 πτ 0 R3 + 3



ωR h

0

2πµ

h3 3 γ dγ = A + Bγ (5.13) ω3

Therefore, we can obtain the rheological parameters through the intercept and slope of the T–γ linear relationship, which can be obtained by Eq. (5.14):

τ0 =

3A 2B (5.14) µ= 3 2πR πR3

where T is the torque, γ is the shear rate, ω is the rotational angular velocity, τ 0 is the yield stress, and μ is the plastic viscosity. A is the T–γ line intercept, and B is the T–γ line slope. The units of yield stress and plastic viscosity are Pa and Pa.s, respectively. Of course, similar rheological parameters can also be obtained through the T–ω relation. However, in some documents (Mendes et al., 2014, Rosquoët et al., 2003, Shamu and Hkansson, 2019), the inhomogeneity of the shear flow at the edge of the parallel plate was reported. Based on the Weissenberg–Rabinowitsch equation, a method to correct the shear stress at the outer edge was obtained. Similarly, it can be corrected by radius, which is r = 2/3R or r = 3/4R. In ASTM C 1749, the viscosity can be expressed by Eq. (5.15):

η=

3hT (5.15)  1d ln T  4 2πR ω  1 + 3d ln ω  

where η is the viscosity in Pa.s, and T is the torque in N.m. 5.2.2.3 Measuring errors and artifacts Similar to the coaxial cylindrical rheometer, the problem of wall slippage is also common in parallel-plate rheometers. Usually, a thin layer appears near the wall of the upper plate and the fluid. Hartman et al. (2002) used ATR spectroscopy to monitor the drop in particle concentration generated in the non-Newtonian fluid near the plate area, and this phenomenon was related to the particle concentration. To reduce the influence of the wall slip phenomenon, many researchers have adopted the surface-roughening method (Alonso et al., 2007, Kalyon and Malik, 2012, Nickerson and Kornfield, 2005, Nehdi and Rahman, 2004). Pawelczyk et al. (2020) and Carotenuto and Minale (2013) believed that the unevenness of the surface would change the actual gap, and they tried to establish a certain correction relationship by expanding the gap. Although the wall slip is not completely eliminated, the corrected viscosity of the Newtonian fluid with rough surface treatment is almost the same as the actual viscosity. Table 5.1 shows the expansion gap under some structural features. Although it is necessary to establish this relationship, its viscosity is unknown for cement paste, and obtaining the correction function will be more complicated due to the shear thickening or thinning behavior. If the measurement results of the rotating rheometer are used as a reference, it is possible to establish the corresponding correction relationship (Ferraris et al., 2001, 2004). The viscosity correction can be obtained by Eq. (5.16) (Pawelczyk et al., 2020, Carotenuto et al., 2012):

ηc =

ηm H (5.16) H +δ

Paste rheometers  103 Table 5.1  Values of δ for different measuring systems at γ  = 5.05 s−1 (Pawelczyk et al., 2020) Effective gap extension δ/mm Measuring system 1 (pyramids) 4 (pyramids) 5 (pyramids) 8 (columns) 12 (bars) 17 (bars)

Silicon oil AK 5000

Silicon oil AK 12,500

25 wt.% suspension (AK 5000)

0.23 ± 0.006 0.91± 0.040 0.44 ± 0.023 0.20 ± 0.003 0.35 ± 0.009 0.73 ± 0.068

0.25 ± 0.004 0.83 ± 0.032 0.49 ± 0.016 0.22 ± 0.009 0.37 ± 0.004 0.71 ± 0.072

022 ± 0.011 0.86 ± 0.071 0.48 ± 0.029 0.19 ± 0.014 035 ± 0.009 0.70 ± 0.089

Figure 5.5 Simulation and data for the PP35-confined geometry. The three lines are for analog data, and the symbols represent experimental data (Ferraris et al., 2007).

where ηc is the corrected viscosity, ηm is the measured viscosity, H is the gap, and δ is the gap expansion value. Some parallel-plate rheometers are tested under the confinement wall, and this may be more advantageous for higher fluidity slurries. In general, the rheological test result under the fence condition is larger compared to that of the plane plate, which may be caused by the friction between the slurry and the wall. Ferraris et al. (2007) improved the parallel-plate rheometer so that it can be used to test cement pastes with larger particle sizes. The experimental data were compared with the numerical simulation results, as shown in Figure 5.5. It can be found that the simulated data under a certain gap has a good correlation with the experimental data. In general, the smaller the gap, the higher the simulation result than the measurement, which may be due to the fact that the fluid outside the upper edge contributes more torque. With Eq. (5.17), the corrected viscosity can be obtained from the measured viscosity:

ηr =

ηm (5.17) h f +1 D

104  Rheology of Fresh Cement-Based Materials

Furthermore, when the gap is much smaller than the diameter of the confinement wall, the corrected viscosity will be closer to the measured viscosity (Ferraris et al., 2007). Compared with the parallel-plate rheometer, the narrow gap coaxial cylindrical rheometer may be more suitable for the rheological measurement of cement slurry, which has some main advantages. For example, it is easier to add samples to the cylinder. Moreover, the outer cylinder can wrap the cement paste, making temperature control effective. By contrast, it is difficult to control the sample under the condition of a parallel plate or cone plate for the slurry with a very strong flowability. Although it can be achieved by setting a boundary wall, this will cause an overestimation of the shear stress. In addition, independent of parallel-plate or cone-plate rheometer, the slurry is susceptible to normal stress, especially for slurries with high viscoelasticity and high thixotropy. Besides, the physical properties of the material can change in the plate rheometer, such as the radial migration and evaporation of water. At present, although many studies are using rotary rheometers to test cement slurries, the rheological parameters of slurries composed of different materials are difficult to compare.

5.2.3 Other rheometers 5.2.3.1 Capillary viscometer 5.2.3.1.1 Geometry Many types of capillary viscometers can be used to measure the viscosity of fluids, especially Newtonian fluids or fluids with Newtonian characteristics. Measurements for cement-based materials usually include pressure and gravity capillary viscometers. The pressure capillary viscometer makes the fluid flow along the capillary tube under constant pressure and obtains the pressure change over a certain distance. The shear stress and shear rate can be determined by the pressure difference and the flow rate to obtain the fluid viscosity. The gravity viscometer records the time it takes for the fluid to pass a certain distance through the capillary tube to determine the fluid viscosity (Demko, 1989). When the liquid flows along the wet pipe wall, its viscosity is proportional to the flow time. The hydrodynamic theory of fluid flowing through a capillary viscometer is very simple and makes necessary assumptions. That is, the fluid flow in the tube is stable, incompressible, and laminar, and hence no slippage occurs at the tube wall (Rosa et al., 2020). The same assumption is also applied to other rheological measurements of cement paste. Capillary rheometers have a high aspect ratio. Normally, the ratio of capillary length to diameter is 30 or even greater. Figure 5.6 shows a commonly used Cannon–Fenske viscometer. The measurement using the capillary viscometer usually contains only one rheological parameter. The viscosity of Newtonian fluids or some cement pastes with properties similar to Newtonian fluids can be obtained through the Poiseuille equation, while the yield stress is still unknown. For pressure capillary viscometers, when the fluid flows in the capillary without plug flow, the pressure drop and flow relationship curve may be used to describe the yield stress and viscosity of the Bingham fluid (Rosa et al., 2020). 5.2.3.1.2 Measurement principle Before measuring, the viscometer is calibrated first. The well-stirred cement slurry was added to the calibrated viscometer and placed in a water bath to keep the temperature constant. Then, the amount of the sample was adjusted until the meniscus was placed 7 mm above the first graduation line. The sample flows freely under its weight, and the time for the meniscus

Paste rheometers  105

Figure 5.6 Cannon–Fenske viscometer (Rosa et al., 2020).

to flow through the first and second graduation lines along the capillary will be measured, with the time accurate to 0.1 s. After that, the capillary was changed to a smaller diameter, and the measurement was repeated. Calculate the average value twice and cannot exceed the required accuracy. The kinematic viscosity of the slurry is determined two times, and the dynamic viscosity can be obtained according to the density. The measurement is conducted concerning ASTM D445-17a. The pressure capillary viscometer needs to measure the pressure and distance at two test points. For gravity capillary viscometers, the viscosity is calculated by Eq. (5.18):

υ = C × t (5.18)

where υ is the kinematic viscosity of the fluid in mm 2 /s, C is the instrument-related constant in mm 2 /s2 , and t is the average flow time in s. The dynamic viscosity is calculated by Eq. (5.19):

η = υ × ρ (5.19)

where η is the dynamic viscosity in mPa.s, and ρ is the fluid density in kg/m3. For pressure capillary viscometer, the shear stress is defined as Eq. (5.20):

τ=

Pr (5.20) 8L

106  Rheology of Fresh Cement-Based Materials

where τ is the shear stress, r is the radius from the center of the capillary, L is the length between two points, and P is the pressure difference. The shear rate at the wall is expressed as Eq. (5.21):

γ =

4Q πR3

(5.21)

where γ is the shear rate, and Q is the flow rate. Dynamic viscosity is obtained by the ratio between shear stress and shear rate as Eq. (5.22):

η=

πR4 P (5.22) 8QL

The relationship between the pressure drop and flow rate is as follows:



Q=

8τ L 16τ 04 L4  πR4  (5.23) P− 0 +  8µ L  3R 3P3 R4 

The shear rate can be corrected by the Rabinowitsch–Mooney equation:

γ =

Q πR3

d ln Q    3 +  (5.24) d ln τ 

5.2.3.1.3 Application to cement paste Capillary viscometers are mainly used in the chemical field, but rarely used for cement slurry. Rosa et al. (2020) used a capillary viscometer for the first time to measure cement slurries with different material compositions and additives, and compared them with the results of a rotary rheometer. The viscosities obtained by the two methods are very close. The test results are shown in Table 5.2. For non-Newtonian fluid suspensions such as cement paste, there are certain restrictions (Mooney, 1931). On the one hand, the diameter of the capillary will be limited by the size of the suspending particles. If the diameter is too small, the capillary will be blocked. If the diameter is too large, the fluid will pass through the pipe quickly, resulting in large errors in readings (Liu et al. 2017). On the other hand, as time passes, the continuous hydration of cement will generate heat, which has a greater influence on the measurement results of the capillary viscometer (Rushing and Hester, 2003, Patterson and Rabouin, 1958). Therefore, a reasonable choice of capillary diameter is a very important prerequisite. ASTM D445-18 provides a standardized method for determining the dynamic and kinematic viscosity of transparent and opaque fluids (Demko, 1989). However, the test method needs to be optimized according to the complexity of cement paste. In summary, capillary viscometer can only provide viscosity parameters related to rheological properties. Throughout the test, the viscosity is time-dependent and only a fixed value can be obtained. If the time and distance of mud movement can be continuously obtained, the change of viscosity can be characterized.

Paste rheometers  107 Table 5.2  Comparison of plastic viscosity between rotational rheometer (RR) and capillary viscometer (C-F)

ηp [Pa.s] Cementitious material CEM I 52.5-SR

CEM II 32.5 BL-II

75% CEM I 52.5-SR + 25% GGBS

75% CEM II 32.5 BL-II + 25% GGBS

w/c

SP/c

φc

RR

C-F 400

0.35 0.35 0.47 0.47 053 0.53 0.63 0.63 0.35 0.35

0.4 0.8 0.4 0.8 0.4 0.8 0.4 0.8 0.4 0.8

0.477 0.477 0.404 0.404 0.376 0.376 0.336 0.336 0.487 0.487

0.123 0.086 0.058 0.047 0.037 0.030 0.019 0.020 0.091 0.055

0.214 0.117 0.087 0.037 0.042 0.022 0.027 0.012 0.114 0.073

0.47

0.4

0.414

0.037

0.030

0.47

0.8

0.414

0.032

0.029

053

0.4

0.386

0.025

0.020

0.63

0.4

0.345

0.013

0.020

0.63

0.8

0.345

0.014

0.013

0.35 0.35

0.4 0.8

0.471 0.471

0.117 0.087

0.169 0.158

0.47

0.4

0.399

0.049

0.058

0.47

0.8

0.399

0.039

0.033

0.53

0.4

0.371

0.034

0.041

0.53

0.8

0.371

0.029

0.026

0.63

0.4

0.332

0.019

0.016

0.63

0.8

0.332

0.019

0.016

0.35 0.35

0.4 0.8

0.476 0.476

0.074 0.067

0.087 0.074

0.47

0.4

0.405

0.035

0.036

0.47

0.8

0.405

0.032

0.028

0.53

0.4

0.377

0.027

0.026

0.53

0.8

0.377

0.019

0.021

0.63

0.4

0.338

0.012

0.022

0.63

0.8

0.338

0.009

0.011

5.2.3.2 Falling sphere viscometer 5.2.3.2.1 Geometry The principle of the falling sphere viscometer is that the speed of the ball falling in the liquid is inversely proportional to the viscosity of the liquid. A typical falling-ball viscometer is shown in Figure 5.7. It usually contains a test tube containing the test liquid and an external tube that controls the temperature. The test tube is slightly inclined, approximately 10°. The viscometer is equipped with spheres of many sizes, ranging from 11.00 to 15.81 mm in diameter. The sphere falls through the liquid in the test tube, and its viscosity is evaluated

108  Rheology of Fresh Cement-Based Materials

Figure 5.7 Typical falling sphere viscometer.

by the time of falling. It is necessary to obtain the calibration coefficient of the instrument through a liquid of known viscosity. The calibration coefficients for different materials and diameters can be found in DIN EN ISO 12058-1. Similar to the capillary viscometer, the falling sphere viscometer also provides only viscosity information that characterizes the rheological behavior of the slurry (Fulmer and Williams, 2002). 5.2.3.2.2 Measurement principle Put the mixed cement paste into the measuring tube, and then put the sphere. The measuring tube is closed by the stopper. Before each series of measurements, the sphere needs to be rolled once along the length of the pipe to make the sphere surface to have complete contact with the cement paste. Turn the viscometer so that the sphere falls from the upper end and passes through the two circular graduation lines in sequence. The time for the upper or lower end of the sphere to pass through the two-scale lines will be measured, and the temperature must be kept constant during the measurement. To reduce the error, at least three measurements are taken. The above test is carried out according to DIN EN ISO 12058-1.

Paste rheometers  109

The dynamic viscosity of the slurry can be obtained by:

η = K ( ρ1 − ρ2 ) t (5.25)

where η is the dynamic viscosity in mPa.s, K is the instrument calibration coefficient in mm 2 /s2 , ρ1 is the density of the sphere in kg/m3, ρ2 is the density of the slurry in kg/m3, and t is the average falling time in s. 5.2.3.2.3 Application to cement paste A falling sphere viscometer is commonly used to test Newtonian fluids and some polymers and resins with low viscosity. For non-Newtonian fluids, there will be more complex flow fields near the sphere, the tube wall and the end region (Munro et al., 1979, Yoshimura, 2000). This results in the measurement of cement slurry being more restricted, and it is difficult to evaluate the shear flow of the sphere in the pipeline. Nicolò et al. (2020) showed that mixtures with different particle sizes lead to the falling process of spheres in fluid showing different shapes, including changes in the trajectory and state of motion. The same effect may also occur in the cement slurry. The measurement of fresh mortar uses a measurement method similar to a falling sphere viscometer, and the viscosity value is obtained according to the Stokes equation (Ferraris, 1999). However, it is difficult to find a similar cement slurry method. In addition, cement slurry is an opaque fluid. Therefore, the falling time of the sphere may be difficult to be accurately measured. Thompson (1949) provided a method to test opaque liquids by the falling sphere method. The realization of induction may have a positive influence on the measurement of cement slurry. The falling sphere viscometer does not have a very solid theoretical basis in the measurement of non-Newtonian fluids, which limits its a­ pplication in the cement slurry. Besides, it can be affected by the complex rheological properties of cement pastes, such as thixotropy, shear thickening, or shear thinning. Therefore, the relationship between viscosity and time is difficult to define. In addition, the variation of the drop motion trajectory and shear rate may be difficult to assess and may require some theoretical support.

5.3 MEASURING PROCEDURES

5.3.1 Flow curves test The well-mixed cement paste is first poured into the outer cylinder of the rheometer. The rotor should be turned gently to place it in the slurry so that the bottom of the rotor is in full contact, avoiding incomplete contact and compaction. Figure 5.8 shows a typical protocol for a flow curve test. During the measurement, it is necessary to maintain the temperature unchanged. By setting the number of shearing steps or the shearing time, the shear rate can be distributed stepwise or continuously. Before the formal measurement, a pre-shearing at the maximum shear rate is required. The shear time usually ranges from 30 to 60 s. Afterward, the sample was set standstill for half a minute to stabilize the torque. The speed quickly reaches the initial speed, and the rising and falling parts will be measured. The rheological parameters are determined by the shear stress and shear rate of the descending section. In addition, the area enclosed by the ascending and the descending curves is positively related to the thixotropy of the material. It can be repeated several times to obtain the average value to improve the accuracy of rheological parameters.

110  Rheology of Fresh Cement-Based Materials

Figure 5.8 The process diagram of flow curve test.

Figure 5.9  Determination of static yield stress in a shear strain test.

For parallel-plate geometry, the added sample should exceed 10%–20% to ensure that the gap is evenly filled by the slurry. Afterward, adjust the gap, and trim off the excess part. Then, the rheometer is run based on the testing protocol.

5.3.2 Static yield stress test The increase in static yield stress depends on the coupling of chemical hydration and interactions between colloidal particles, van der Waals forces, etc. With elapsed time, a denser network structure is formed between the particles, which increases the yield stress. The slurry is first pre-sheared to reduce the effects of thixotropy. After that, the slurry was applied with a shear rate (or shear strain) of 10 –1 to 10 –3 s−1, and the shear stress at the critical state was the static yield stress (Feys et al., 2018). This critical state corresponds to the transition from rising to falling shear stress during shearing. Figure 5.9 shows the determination of the static yield stress and critical strain.

Paste rheometers  111

Figure 5.10 Evolution of the storage and loss moduli as a function of the stress amplitude.

5.3.3 Oscillatory shear test 5.3.3.1 Description of SAOS and LAOS Small-amplitude oscillatory shear (SAOS) is a technical method to evaluate the deformation ability of fluid based on the classic Hooke’s law. In SAOS, a strain smaller than the critical strain of slurry is applied for a continuous sinusoidal excitation to obtain a stress response, such that the investigated slurry is deformed only within its elastic region. However, the stress response usually has a certain degree of lag (Théau et al., 2016). Once the strain exceeds a critical value, the material structure is destroyed, and the stress–strain relationship is no longer linear (Betioli et al., 2009), as shown in Figure 5.10. For polymers and suspensions, the critical value of structural transformation can be determined (Yuan et al., 2017, Betioli et al., 2009, Nachbaur et al., 2001, Kallus et al., 2001). Therefore, the SAOS test method identifies the process of materials from construction to destruction. The association of large-amplitude oscillatory shear (LAOS) data with rheological behavior provides a method for evaluating fluid rheological properties (Cho et al., 2005). Compared with SAOS, the slurry is subjected to a strain amplitude higher than the critical strain. The stress–strain relationship is no longer a pure linear relationship, and the stress cannot be expressed by a single trigonometric function. In the case of nonlinearity, the modulus is no longer independent of the strain amplitude, so there will be periodic deviations (Hyun et al., 2011). High-order harmonics have a certain contribution to affecting the behavior of non-Newtonian fluids (Giacomin et al., 2011). Fourier transform is considered to be the most commonly used technical means to describe the properties of nonlinear behavior fluids (Hyun et al., 2002). Hyun et al. (2011) developed Fourier transform based on nuclear magnetic resonance spectroscopy and developed the Fourier transform–rheology method. The method has been applied to rheological tests of various complicated fluids. Hyun et al. (2002) proposed the concepts of generalized storage modulus and generalized loss modulus, and used them as a measure to evaluate the nonlinear behavior of viscoelasticity. Corresponding to the Fourier transform method, a higher accuracy is obtained. In a later report, the theoretical development and application of LAOS data interpretation were reviewed (Hyun et al., 2011). In addition, through Lissajous curves, the complicated high-order harmonic problem is transformed into a mathematical geometric problem, and the slope of a certain point of the stress–strain curve is used as the physical interpretation of the LAOS data (Simon, 2018). For SAOS, the ratio of shear stress to strain can be expressed by the complex modulus, including the storage modulus and loss modulus. Among them, the storage modulus

112  Rheology of Fresh Cement-Based Materials

represents the amount of energy stored in the slurry due to deformation under elastic behavior, and the loss modulus represents the amount of heat or other losses that need to be overcome in the process, such as resistance between particles. LAOS also includes two parameters representing elastic and viscous behaviors, and the dynamic modulus related to the material is obtained through Fourier transformation. If the contribution of higher harmonics is ignored, the storage modulus and loss modulus under LAOS conditions may lose their meanings. However, it may still provide some information about microstructure changes to distinguish complex fluids (Hyun et al., 2002). LAOS is mainly carried out under strain or stress control through the Fourier transform method; that is, the output stress obtained is decomposed into a Fourier series. Currently, the most commonly used method for interpreting LAOS data is the geometric method of Lissajous curves. 5.3.3.2 Measurement principle For SAOS, the strain can be obtained by Eq. (5.26):

γ (t) = γ 0 × sin(ω t) (5.26)

where γ(t) is the strain at time t, γ0 is the maximum amplitude, and t is the time. Then, the complex modulus can be obtained:

G* = τ / γ (5.27)



G* = G′ + iG′′ (5.28)

The relationship among the rheological parameters can be expressed in Eqs. (5.29–5.31):

G′ =

τ0 × sin δ (5.29) γ0



G′′ =

τ0 × cos δ (5.30) γ0



tan δ =

G′′ (5.31) G′

where G′ is the storage modulus, G″ is the loss modulus, τ0 is the maximum stress, γ0 is the maximum strain, and δ is the phase angle. For purely elastic materials, the phase angle is 0°, and the response is completely flexible. For pure Newtonian fluids, the phase angle is 90°, and the reaction is completely vicious. For cement-based materials, the response is somewhere in between, showing viscoelasticity that corresponds to the elastic and viscous behaviors of the material. SAOS is mainly performed using the abovementioned rotational rheometer. When it is a strain-controlled rheometer, SAOS technique usually includes the following three processes: Strain sweep. In this test, the frequency needs to be kept at a certain value (generally 1~2 Hz for cement paste). The material receives a continuous sinusoidal strain, and the amplitude gradually increases, usually from 10 –5% to 1%. When the strain exceeds a certain critical value, the internal structure breaks down, and the modulus decreases as the

Paste rheometers  113

Figure 5.11 Typical oscillatory strain sweep of fresh cement pastes at 1 Hz. (a) Critical value can be easily recognized, and (b) critical value is difficult to recognize.

strain increases. Through this process, the linear viscoelastic region (LVER) and the critical strain of the material can be obtained. In some cases, the critical strain is easy to identify (see Figure 5.11a). However, the critical strain may not be easily determined in other cases (see Figure 5.11b). Oscillatory frequency sweep: Oscillatory frequency sweep was used to evaluate the stability of the material over a wide frequency range. The amplitude was kept constant, while the frequency was increased incrementally from 0.01 to 100 Hz. Both storage modulus and loss modulus were monitored during the tests. Oscillatory time sweep: Oscillatory time sweep test was used to investigate the evolution of rheological properties due to structural changes. In this test procedure, the amplitude was kept constant at a magnitude lower than the critical strain, and the frequency was also kept constant. Generally, the applicable frequency of cement paste is 1 Hz for LAOS. The process carried out is very similar to the abovementioned SAOS process. Taking strain input as an example, it includes three processes: strain oscillation sweep, frequency oscillation sweep, and time oscillation sweep. Compared with SAOS, LAOS imposes larger stress or strain amplitude without a large difference in frequency. The main purpose is to analyze the transition of the stress–strain curve from the linear region to the nonlinear region. If the frequency is too high or too low, it will affect the rheological test. At higher frequencies, it may be affected by the inertia of the instrument and fluid. At the same time, for time-dependent fluids, the development and evolution of the internal structure of the material make the lower LAOS parameter at frequency relatively greatly affected. When using a strain-controlled rheometer, the input stress is as Eq. (5.32) (Läuger and Stettin, 2010, Théau et al., 2016):

τ (t) = τ 1 × sin(ω t) (5.32)

Due to symmetry, the Fourier series of the output strain is expressed as Eq. (5.33):

γ (t) =

∑γ

n = odd

n

sin ( nω t − δ n ) (5.33)

114  Rheology of Fresh Cement-Based Materials

Strain can be divided into elastic and viscous parts, where Jn′ and Jn′′ are the storage modulus and loss modulus, respectively:

γ (t) = τ 1

∑ [ J ′ sin(nω t) − J ′′ cos(nω t)] (5.34) n

n

n= odd

The strain rate can also be expressed by material-related fluid properties ψ n′ and ψ n′′:

γ (t) = τ 1

∑ [ nψ ′′ cos(nω t) + nψ ′ sin(nω t)] (5.35) n

n

n = odd

Furthermore, small-strain compliance and large-strain compliance are defined, corresponding to the slope of the stress–strain curve in Lissajous curves:

JM ′ =



JM ′ =

dγ = JM ′ (ω : τ 1 ) (5.36) dτ τ =0

∑ nJ ′ (5.37) n

n=odd



JL′ =



JL′ =

γ

= JL′ (ω : τ 1 ) (5.38)

τ τ =±τ1

∑ J ′ (5.39) n

n =odd

Therefore, the stress-softening ratio R can be represented by Eq. (5.40):

R (ω : τ 1 ) =

JL′ − JM ′ (5.40) JL′

Similarly, the definition of small rate fluidities and large rate fluidities corresponds to the slope of the stress–strain rate curve in Lissajous curves. dγ = JM ′ (ω : τ 1 ) (5.41) dτ τ =0



ψM ′ =



ψM ′ =ω

∑ n J ′′ (5.42) 2

n

n= odd

γ



ψ L′ =



ψ L′ = ω

τ τ =±τ1

= ψ L′ (ω : τ 1 ) (5.43)

∑ nJ ′′ (5.44) n

n=odd

Paste rheometers  115

The shear-thinning ratio Q is defined as Eq. (5.39):

Q (ω , τ 1 ) =

ψ L′ − ψ M ′ (5.45) ψ L′

In particular, when R = 0, it should correspond to the critical strain of SAOS. At the same time, different R and Q values correspond to different shapes of Lissajous curves. Therefore, the LAOS data can be explained by changes in the geometric relationship between the stress and strain or strain rate curves. 5.3.3.3 Application to cement paste SAOS is widely used in high-molecular-weight polymers, suspensions, emulsions, greases, etc., and it can also be used in cement paste to evaluate the viscoelasticity and the evolution of internal structure (Yuan et al., 2017, Betioli et al., 2009, Nachbaur et al., 2001, (Papo and Caufin, 1991)). Schultz and Struble (1993) pointed out that the storage modulus of cement slurry is approximately 14–24 kPa and the critical strain is approximately 10 –4. As w/c increases, the storage modulus and critical strain decrease slightly. However, Yuan et al. (2017) believed that the critical strain is between 10 –5 and 10 –4. The value is significantly affected by the use of a water-reducing agent, while it shows a weak correlation with w/c. Researchers have gradually connected the rheological parameters of SAOS with the structural development of cement paste and even the formation of hydration products on a microscopic scale. Betioli et al. (2009) found that SAOS parameters can correspond to the heat of the hydration process. Huang et al. (2020) established a linear relationship between the storage modulus and ettringite (AFt) content. Figure 5.12 shows the correlation between ettringite content and storage modulus. These results also prove that SAOS may have a strong correlation between early hydration progress and structure establishment. In addition, Yuan et al. (2017) found that the storage modulus was in good agreement with the static yield stress growth, as shown in Figure 5.13. The yield stress can be expressed as the product of the critical strain and the corresponding modulus (Betioli et al., 2009). Ukrainczyk et al. (2020) compared the yield stress calculated

Figure 5.12  Correlation of the ettringite content (wt.%) and the storage modulus of C3A-gypsum paste (Huang et al., 2020).

116  Rheology of Fresh Cement-Based Materials

Figure 5.13 Comparison of yield stress (a) and storage modulus (b) over time (Yuan et al., 2017).

by SAOS and the mechanical model, and verified the relationship between the yield stress and SAOS parameters. Sun et al. (2006) used a shear wave reflection technique to monitor the early behavior of cement slurries with different water–cement ratios. The results have a good correlation with the storage modulus obtained directly using SAOS. The small deformation under SAOS may not fully characterize the processing and molding of some materials. In this context, LAOS corresponding to the study of rheology under large deformation or deformation rates, mainly for the measurement of polymers (Hyun et al., 2002), is used to measure the structural behavior of some fluids with high viscosity or deformation ability. However, the research of LAOS on the measurement of the rheological properties of cement pastes is limited. Théau et al. (2016) first used LAOS to study the influence of amplitude and frequency on the rheological parameters of cement slurries, and the underlying mechanisms were explained by a Lissajous–Bowditch curve. Stress–strain or stress–strain rate curves or geometric properties of stress, such as shape, slope, enclosed area, and evolution, were linked to the thixotropy and viscoelastic behavior of the material (Qian et al., 2019). Figure 5.14 shows the LAOS parameters of cement paste.

Figure 5.14 A ge evolution of the nonlinear viscoelastic parameters of the NC cement paste at 1 Hz and 40 Pa; R: stress-softening ratio; Q: shear-thinning ratio (Théau et al., 2016).

Paste rheometers  117

5.4 SUMMARY This chapter reviews the measurement of the rheological properties of cement paste, mainly including rheological equipment, measuring principles, testing methods, and data processing. The following conclusions can be reached. Surface-roughening methods include sandblasting, grooving, recessing or protruding, and even blade rotors, which have a certain restrictive effect on the occurrence of wall slip. However, the blade rotors may cause complicated flow. The geometric height of the roughstructure design will also affect its restrictive effect on wall slip. The coaxial cylindrical rheometer has better temperature control ability for cement slurry samples and ensures its uniformity. Compared with the parallel-plate rheometer, the cement slurry overflow from the edge is avoided. It is more suitable for the measurement of the rheological properties of most cement pastes. The storage modulus obtained by SAOS has good consistency with the yield stress change of cement slurry. SAOS is a promising technique for monitoring the development of microstructure in real time. For LAOS with larger strain, the Lissajous curve is the most suitable to interpret the data, and its shape evolution and LAOS rheological parameters can be well connected. The viscosities measured by the capillary viscometer and the rotational rheometer are similar. However, it is difficult to characterize its rheology through two or even three parameters of the existing rheology model. Falling sphere viscometers are difficult to apply to cement paste because of the complex flow behavior. The rheometer can quickly and accurately evaluate the rheological properties of cement paste. Therefore, it has been widely used in infrastructure construction and experimental research. However, there is no clear specification for evaluating the applicability of each rheometer and the quantitative relationship between rheological parameters. More research is needed on this aspect. REFERENCES Alderman, N. J., Meeten, G. H., and Sherwood, J. D. (1991). “Vane rheometry of bentonite gels.” Journal of Nonnewtonian Fluid Mechanics, 39(3), 291–310. Alonso, M. M., Palacios, M., Puertas, F., et al. (2007). “Effect of polycarboxylate admixture structure on cement paste rheology.” Materiales De Construccion, 57(286), 65–81. ASTM C 1749-17a. Standard guide for measurement of the rheological properties of hydraulic cementious paste using a rotational rheometer. American Society for Testing and Materials, 2017. ASTM D 445-18. Standard Test Method for Kinematic Viscosity of Transparent and Opaque Liquids (and Calculation of Dynamic Viscosity). Barnes, H. A. (1990). “The vane‐in‐cup as a novel rheometer geometry for shear thinning and thixotropic materials.” Journal of Rheology, 34(6), 841–866. Betioli, A. M., Gleize, P. J. P., Silva, D. A., et al. (2009). “Effect of HMEC on the consolidation of cement pastes: Isothermal calorimetry versus oscillatory rheometry.” Cement & Concrete Research, 39(5), 440–445. Cardoso, F. A., Fujii, A. L., Pileggi, R. G., and Chaouche, M. (2015). “Parallel-plate rotational rheometry of cement paste: Influence of the squeeze velocity during gap positioning.” Cement & Concrete Research, 75, 66–74. Carotenuto, C., and Minale, M. (2013). “On the use of rough geometries in rheometry.” Journal of Non-Newtonian Fluid Mechanics, 198, 39–47. Carotenuto, C., Marinello, F., and Minale, M. (2012). “A new experimental technique to study the flow in a porous layer via rheological tests.” AIP Conference Proceedings, 1453(1), 29–34.

118  Rheology of Fresh Cement-Based Materials Cepuritis, R., Skare, E. L., Ramenskiy, E., et al. (2019). “Analyzing limitations of the FlowCyl as a one-point viscometer test for cement paste.” Construction and Building Materials, 218, 333–340. Chen, S., and Rothstein, J. P. (2004). “Flow of a wormlike micelle solution past a falling sphere.” Journal of Non-Newtonian Fluid Mechanics, 116(2–3), 205–234. Cho, K. S., Hyun, K., Ahn, K. H., et al. (2005). “A geometrical interpretation of large amplitude oscillatory shear response.” Journal of Rheology, 49(3), 747–758. DIN EN ISO 12058-1. Determination of viscosity using a falling-ball viscometer-Part 1: Inclinedtube method. Demko, J. M. (1989). “Development of an ASTM standard test method for measuring engine oil viscosity using capillary viscometers at high-temperature and high-shear rates.” ASTM Special Technical Publication, 1068, 11. Deng, D., Zhu, R., Peng, J., et al. (2013). “Effect of superplasticizers and limestone powders on shear thickening behavior of cement paste.” Journal of Building Materials, 16(5), 744–752. Erdem, T. K., Ahari, R. S., et al. (2015). “Thixotropy and structural breakdown properties of self consolidating concrete containing various supplementary cementitious materials.” Cement & Concrete Composites, 59, 26–37. Ferraris, C. F. (1999). “Measurement of the rheological properties of cement paste: A new approach.” Journal of Research of the National Institute of Standards and Technology, 104(5), 333–342. Ferraris, C. F., Beaupr, D., et al. Comparison of concrete rheometers: International tests at MB (Cleveland OH, USA) in May, 2003. US Department of Commerce, National Institute of Standards and Technology, 2004. Ferraris, C. F., Brower, L. E., Banfill, P., et al. Comparison of concrete rheometers: International test at LCPC (Nantes, France) in October, 2000. US Department of Commerce, National Institute of Standards and Technology, 2001. Ferraris, C. F, and Gaidis, J. M. (1992). “Connection between the rheology of concrete and rheology of cement paste.” ACI Materials Journal, 89(4), 388–393. Ferraris, C. F., Geiker, M., Martys, N. S., and Muzzatti, N. (2007). “Parallel-plate rheometer calibration using oil and computer simulation.” Journal of Advanced Concrete Technology, 5(3), 363–371. Ferraris, C. F., Martys, N. S., and George, W. L. (2014). “Development of standard reference materials for rheological measurements of cement-based materials.” Cement and Concrete Composites, 54, 29–33. Ferraris, C. F., Obla, K. H., and Hill, R. (2001). “The influence of mineral admixtures on the rheology of cement paste and concrete.” Cement & Concrete Research, 31(2), 245–255. Feys, D., Cepuritis, R, Jacobsen, S., et al. (2018). “Measuring rheological properties of cement pastes: Most common techniques, procedures and challenges.” RILEM Technical Letters, 2, 129–135. Fulmer, E. I., and Williams, J. C. (2002). “A method for the determination of the wall correction for the falling sphere viscometer.” Journal of Physical Chemistry, 40(1), 143–149. Giacomin, A. J., Bird, R. B., Johnson, L. M., et al. (2011). Large-amplitude oscillatory shear flow from the corotational Maxwell model. Journal of Non-Newtonian Fluid Mechanics, 166, 1081–1099. Glab, C., Wca, B., Lca, B., et al. (2020). “Rheological properties of fresh concrete and its application on shotcrete - ScienceDirect.” Construction and Building Materials, 243, 118180. Guoju, K., Zhang, J., Xie, S., et al. (2020). “Rheological behavior of calcium sulfoaluminate cement paste with supplementary cementitious materials.” Construction and Building Materials, 243, 118234. Haimoni, A., and Hannant, D. J. (1988). “Developments in the shear vane test to measure the gel strength of oilwell cement slurry.” Advances in Cement Research, 1(4), 221–229. Hanehara, S., and Yamada, K. (1999). “Interaction between cement and chemical admixture from the point of cement hydration, absorption behavior of admixture, and paste rheology.” Cement & Concrete Research, 29(8), 1159–1165. Harboe, S, Modigell, M., et al. (2011). “Wall slip of semi-solid A356 in Couette rheometers.” AIP Conference Proceedings, 1353(1), 1075–1080.

Paste rheometers  119 Harboe, S., Modigell, M., and Pola, A. (2012). “Wall slip effect in Couette rheometers.” International Conference on Semisolid Processing of Alloys and Composites S2P 2012, 192, 353–358. Hartman Kok, P. J. A., Kazarian, S. G., Lawrence, C. J., et al. (2002). “Near-wall particle depletion in a flowing colloidal suspension.” Journal of Rheology, 46(2), 481–493. Hu, C, de Larrard, F., Sedran, T., et al. (1996). “Validation of BTRHEOM, the new rheometer for soft-to-fluid concrete.” Materials & Structures, 29(10), 620–631. Huang, T., Yuan, Q., and He, F., et al. (2020). “Understanding the mechanisms behind the timedependent viscoelasticity of fresh C 3 A-gypsum paste.” Cement and Concrete Research, 133, 106084. Hyun, K, Kim, S. H., Ahn, K. H., et al. (2002). “Large amplitude oscillatory shear as a way to classify the complex fluids.” Journal of Non-Newtonian Fluid Mechanics, 107(1–3), 51–65. Hyun, K, Wilhelm, M, Klein, C. O., et al. (2011). “A review of nonlinear oscillatory shear tests: Analysis and application of large amplitude oscillatory shear (LAOS).” Progress in Polymerence, 36(12), 1697–1753. Kallus, S., Willenbacher, N., Kirsch, S., et al. (2001). “Characterization of polymer dispersions by Fourier transform rheology.” Rheologica Acta, 40(6), 552–559. Kalyon, D. M, and Malik, M. (2012). “Axial laminar flow of viscoplastic fluids in a concentric annulus subject to wall slip.” Rheologica Acta, 51(9), 805–820. Keentok, M., Milthorpe, J. F, and O’Donovan, E. (1985). “On the shearing zone around rotating vanes in plastic liquids: Theory and experiment.” Journal of Non-Newtonian Fluid Mechanics, 17(1), 23–35. Khayat, K. H. (1998). “Viscosity-enhancing admixtures for cement-based materials — An overview.” Cement and Concrete Composites, 20(2–3), 171–188. Koehler, E, and Fowler, D. (2004). “Development of a portable rheometer for fresh portland cement concrete.” Technical Reports, International Center for Aggregates Research, The University of Texas at Austin, Austin, United States. Läuger, J., and Stettin, H. (2010). “Differences between stress and strain control in the nonlinear behavior of complex fluids.” Rheologica Acta, 49(9), 909–930. Lewis, J. A., Matsuyama, H., Kirby, G., et al. (2000). “Polyelectrolyte effects on the rheological properties of concentrated cement suspensions.” Journal of the American Ceramic Society, 83(8), 1905–1913. Liu, Y., Shi, C., Jiao, D., and An, X. (2017). “Rheological properties, models and measurements for fresh cementitious materials-a short review.” Journal of the Chinese Ceramic Society, 45(05), 708–716. Liu, Y., Shi, C., Yuan, Q., et al. (2020). “The rotation speed-torque transformation equation of the Robertson-Stiff model in wide gap coaxial cylinders rheometer and its applications for fresh concrete.” Cement and Concrete Composites, 107, 103511. Ma, K., Feng, J., Long, G., et al. (2016). “Effects of mineral admixtures on shear thickening of cement paste.” Construction & Building Materials, 126, 609–616. Maybury, J., Ho, J. C. M., and Binhowimal, S. A. M. (2017). “Fillers to lessen shear thickening of cement powder paste.” Construction & Building Materials, 142, 268–279. Mca, B, Lei, Y, Yan, Z. A., et al. (2020). “Yield stress and thixotropy control of 3D-printed calcium sulfoaluminate cement composites with metakaolin related to structural build-up.” Construction and Building Materials, 252, 119090. Mendes, P. R. D. S., Alicke, A. A., and Thompson, R. L. (2014). “Parallel-plate geometry correction for transient rheometric experiments.” Applied Rheology, 24(5), 52721. Mooney, M. (1931). “Explicit formulas for slip and fluidity.” Journal of Rheology, 2(2), 210. Morinaga, S. (1973). “Pumpability of concrete and pumping pressure in pipelines.” Proceedings of Rilem Seminar, Leeds, 3, 1–39. Munro, R. G., Piermarini, G. J., Block, S. (1979). “Wall effects in a diamond‐anvil pressure‐cell falling‐sphere viscometer.” Journal of Applied Physics, 50(5), 3180–3184. Nachbaur, L., Mutin, J. C., Nonat, A., et al. (2001). “Dynamic mode rheology of cement and tricalcium silicate pastes from mixing to setting.” Cement & Concrete Research, 31(2), 183–192.

120  Rheology of Fresh Cement-Based Materials Nehdi, M, and Rahman, M. A. (2004). “Estimating rheological properties of cement pastes using various rheological models for different test geometry, gap and surface friction.” Cement & Concrete Research, 34(11), 1993–2007. Nickerson, C. S., and Kornfield, J. A. (2005). “A “cleat” geometry for suppressing wall slip.” Journal of Rheology, 49(4), 865–874. Nicolò, R. S., Davaille, A., Kumagai, I., et al. (2020). “Interaction between a falling sphere and the structure of a non-Newtonian yield-stress fluid.” Journal of Non-Newtonian Fluid Mechanics, 284, 104355. Ortega-Avila, J. F., Pérez-González, J., Marín-Santibáñez, B. M., et al. (2016). “Axial annular flow of a viscoplastic microgel with wall slip.” Journal of Rheology, 60(3), 503–515. Papo, A., and Caufin, B. (1991). “A study of the hydration process of cement pastes by means of oscillatory rheological techniques.” Cement and Concrete Research, 21(6), 1111–1117. Patterson, G. D., and Rabouin, L. H. (1958). “Capillary viscometer for high-temperature measurements of polymer solutions.” Review of Scientific Instruments, 29(12), 1086–1088. Pawelczyk, S., Kniepkamp, M., Jesinghausen, S., et al. (2020). “Absolute rheological measurements of model suspensions: Influence and correction of wall slip prevention measures.” Materials, 13(2), 467. Proske, T., Rezvani, M., and Graubner, C. A. (2020). “A new test method to characterize the pressure-dependent shear behavior of fresh concrete.” Construction and Building Materials, 233, 117255. Qian, Y., Ma, S. W., Kawashima, S., and Schutter, G. D. (2019). “Rheological characterization of the viscoelastic solid-like properties of fresh cement pastes with nanoclay addition.” Theoretical and Applied Fracture Mechanics, 103, 102262. Rosa, N. D. L., Poveda, E., Ruiz, G., et al. (2020). “Determination of the plastic viscosity of superplasticized cement pastes through capillary viscometers.” Construction and Building Materials, 260, 119715. Rosquoët, F., Alexis, A., Khelidj, A., et al. (2003). “Experimental study of cement grout: Rheological behavior and sedimentation.” Cement and Concrete Research, 33(5), 713–722. Roussel, N., Ovarlez, G., Garrault, S., et al. (2012). “The origins of thixotropy of fresh cement pastes.” Cement and Concrete Research, 42(1), 148–157. Rushing, T. S., and Hester, R. D. (2003). “Low-shear-rate capillary viscometer for polymer solution intrinsic viscosity determination at varying temperatures.” Review of Scientific Instruments, 74(1), 176–181. Saak, A. W., Jennings, H. M., and Shah, S. P. (2001). “The influence of wall slip on yield stress and viscoelastic measurements of cement paste.” Cement & Concrete Research, 31(2), 205–212. Schüller, R. B., and Salas-Bringas, C. (2007). “Fluid temperature control in rotational rheometers with plate-plate measuring systems.” Psychologie, 15, 159–163. Schultz, M. A., and Struble, L. J. (1993). “Use of oscillatory shear to study flow behavior of fresh cement paste.” Cement and Concrete Research, 23(2), 273–282. Shamu, T. J., and Hkansson, U. (2019). “Rheology of cement grouts: On the critical shear rate and no-slip regime in the Couette geometry.” Cement and Concrete Research, 123, 105769. Simon, R. (2018). “Large amplitude oscillatory shear: Simple to describe, hard to interpret.” Physics Today, 71(7), 34–40. Sun, B. J., Gao, Y. H., and Liu, D. Q. (2007). “Experimental study on rheological property for cement slurry and numerical simulation on its annulus flow.” Journal of Hydrodynamics (Ser.A), 22(3), 317–324. Sun, Z., Voigt, T., and Shah, S. P. (2006). “Rheometric and ultrasonic investigations of viscoelastic properties of fresh Portland cement pastes.” Cement & Concrete Research, 36(2), 278–287. Théau, C., and Mohend, C., et al. (2016). “Rheological behavior of cement pastes under large amplitude oscillatory shear.” Cement & Concrete Research, 89, 332–344. Thompson, A. M. (1949). “A falling-sphere viscometer for use with opaque liquids.” Journal of Scientific Instruments, 26(3), 75. Ukrainczyk, N., Thiedeitz, M., Krnkel, T., et al. (2020). “Modeling SAOS yield stress of cement suspensions: Microstructure-based computational approach.” Materials, 13(12), 2769.

Paste rheometers  121 Wallevik, J. E. (2009). “Rheological properties of cement paste: Thixotropic behavior and structural breakdown.” Cement & Concrete Research, 39(1), 14–29. Wang, W., Zhu, H., De Kee, D., et al. (2010). “Numerical investigation of the reduction of wallslip effects for yield stress fluids in a double concentric cylinder rheometer with slotted rotor.” Journal of Rheology, 54(6), 1267–1283. Yoshimura, A. (2000). “Wall slip corrections for Couette and parallel disk viscometers.” Journal of Rheology, 32(1), 53–67. Yang, H., Lu, C., and Mei, G. (2018). “Shear-thickening behavior of cement pastes under combined effects of mineral admixture and time.” Journal of Materials in Civil Engineering, 30(2), 04017282. Yoshimura, A. S., and Prud’homme, R. K. (1988). “Viscosity measurements in the presence of wall slip in capillary, Couette, and parallel-disk geometries. SPE Reservoir Engineering, 3(02), 735–742. Yuan, Q, Lu, X, Khayat, K. H., et al. (2017). “Small amplitude oscillatory shear technique to evaluate structural build-up of cement paste.” Materials & Structures, 50(2), 112. Zhu, H., Martys, N. S., Ferraris, C., et al. (2010). “A numerical study of the flow of Bingham-like fluids in two-dimensional vane and cylinder rheometers using a smoothed particle hydrodynamics (SPH) based method.” Journal of Non-Newtonian Fluid Mechanics, 165(7–8), 362–375.

Chapter 6

Concrete rheometers

6.1 INTRODUCTION For a long time in the past, the workability or, more precisely, flow property of fresh conventional concrete has been tested predominately by the slump test (C143, 2008). However, the applications of self-compacting concrete (SCC), high-performance concrete (HPC), high flowable concrete, and other types of concrete introduce a wide range of materials into engineering practice. The compositional complexity of concrete mix makes its flow behavior very sensitive to slight changes in mixture proportions. The slump test becomes a less reliable indicator of the workability of fresh concrete. So far, approximately 100 tests have been developed to measure concrete flow over the past few decades (Roussel, 2011). These tests fall into the empirical approach and the scientific approach. The empirical tests have been described in Chapter 4. The scientific approach describes the material itself and can understand the intrinsic properties of fresh concrete, which is called rheology. Through rheology, the flow and deformation property (rheological property) of fresh concrete can be defined strictly in terms of physical constants with the fundamental unit, and basic rheological principles can guide the study of the physical and analytical models of the material. During the 1970s, measuring apparatus with rotating vanes or coaxial cylinders had been used to make the theoretical analysis of the flow behavior of fresh concrete (Banfill, 2006). Based on the two-point test, Tattersall and Banfill (1983) developed a method for measuring the power requirements during the mix. It can be deployed both in the lab and on-site, standing for a significant step forward to characterize the flow properties of fresh concrete. The two-point test measures the values of shear stress under a minimum of two shear rates, and then calculates the rheological parameters of the material. In the past few decades, a large variety of materials and admixtures, together with many new processing methods, have been introduced into the field of concrete. The rheology of fresh concrete has been comprehensively studied, and its applications are greatly extended. It is possible for us to predict the fresh properties, design and select materials, and achieve the required concrete performance. According to the US National Institute of Standards and Technology (NIST), all the flow tests can be classified into four groups (Hackley and Ferraris, 2001): confined flow tests, free flow tests, vibration tests, and rotational rheometers. The first three types belong to empirical tests. The use of rotational rheometers is the only method that can determine the exact rheological properties of a test sample in a fundamental rheological unit. This chapter introduces several prototypes of rotational rheometers for concrete. The basic principles and measurement procedures are introduced, and errors and artifacts are discussed. Finally, the relations of rheological parameters measured by different rheometers are summarized. DOI: 10.1201/9781003265313-6

123

124  Rheology of Fresh Cement-Based Materials

6.2 TESTS METHODS AND PRINCIPLES There are several rheometry methods, including rotational method, capillary method, drainage vessel method, and oscillation method. The most commercially available rheometers for fresh concrete are the rotational rheometers. Rotational rheometry can be classified regarding geometrical design and mode of operation. Four geometries used in rotational rheometers are coaxial, parallel plates, cone-plate, and impeller (See Figure 6.1). In a coaxial rheometer, the fluid is placed in a cylindrical cup, and a coaxial but smaller cylinder is submerged in the fluid. In a parallel-plate rheometer, two disks are positioned parallel to each other. Moreover, the impeller rheometers are based on rotating an impeller, which has various shapes. The test sample is placed between disks. The material is sheared as one disk rotates, while the other remains stationary. Compared to the parallel-plate rheometer, the cone-plate rheometer replaces one plate with a cone. However, since the coarse aggregates at the bottom of the container often block the immersion of the cone, this type of rheometer is rarely used for fresh concrete. Several concrete rheometers are shown in Figure 6.2. The ICAR rheometer and BML Viscometer are coaxial cylinder types. The BTRHEOM is parallel-plate type, and the Tattersall two-point and IBB rheometers are impeller type. The Tattersall two-point rheometer (MK II) is one of the first instruments to use impeller geometry. It is also the earliest attempt to measure the rheological properties of fresh concrete with the Bingham model. The Tattersall two-point rheometer has two different impeller types, which allow measuring a wide range of concrete mixtures. IBB is, in fact, the automated version of MK III, which is the modified model of the two-point rheometer (Beaupre, 1994). BML Viscometer and ConTec are similar devices. Viscometer 5 is heavy and suitable for laboratory research, while the ICAR and BTRHEOM can be used in the laboratory or on-site. In the rest of this section, three typical concrete rheometers are introduced. The basic information including their geometries, principles, and other related issues will be described in detail and compared.

Figure 6.1 Typical rheometer geometry configurations: (a) coaxial cylinder, (b) parallel plate, (c) cone-plate, and (d) impeller.

Concrete rheometers  125

Figure 6.2  Concrete rheometers (Brower and Ferraris, 2001): (a) Tattersall two-point rheometer, (b) IBB rheometer, (c) ICAR rheometer, (d) BTRHEOM rheometer, and (e) BML Viscometer.

6.2.1 Coaxial cylinder rheometer Two basic operation modes, i.e., stress-controlled and rate-controlled modes, are available to convert the above geometries into a rotational rheometer. The stress-controlled mode measures the resulting shear rate by controlling the stress input, while the rate-controlled model measures the resulting shear stress through controlling shear rate input (Schramm, 1994). Some rheometers can work in both test modes, while most commercial concrete rheometers usually adopt the rate-controlled mode. Operationally, either the inner or the outer cylinder is rotated with the counterpart (the other cylinder) fixed. For the rheometers with the inner cylinder rotating, i.e., Searle-type rheometers, the torque is measured at the inner cylinder. For the rheometers with the outer cylinder rotating, i.e., Couette type, the inner cylinder is freely suspended from a torsion wire, the resistance of the flow causes wire deflections on the inner cylinder, and then the torque is recorded. Moreover, both the drive on the rotor (inner cylinder) and the torque detector of Searle-type rheometers act on the same rotating axis and may lower the measuring accuracy compared to Couette type. However, as technology advances, the differences between the Searle-type and the Couette-type rheometers are no longer significant. The inner cylinder of concrete rheometers is generally substituted by a vane or ribs to facilitate the placement of fresh concrete. The ICAR rheometer is a typical commercial Searle rheometer, while the ConTec Viscometer 5 is a typical Couette rheometer. The measurement procedure involves increasing and decreasing the speed (in preset discrete increments) while measuring the torque at each speed. The problem is to relate speed and torque to shear rate and shear stress to calculate the exact rheological properties in fundamental units. Therefore, a transformation equation should be derived so that the regression analysis can acquire the parameters of a given rheological model. Compared to the radius of the inner cylinder, the width of the annulus for most of the general-purpose rheometer is relatively narrow. Therefore, calculating the shear rate and shear stress using an average radius will not introduce too many errors. However, these narrow-gaped rheometers do not apply to concrete. The maximum aggregate size should be considered for the concrete rheometer. Studies showed that the minimum gap size should be at least three times the aggregate size so that the inhomogeneity of the material can be eliminated to a large extent during the test (Koehler and Fowler, 2004).

126  Rheology of Fresh Cement-Based Materials

Figure 6.3 Components of ICAR rheometer.

6.2.1.1 Searle rheometer The ICAR rheometer is a mobile, portable rheological test device for fresh concrete, as shown in Figure 6.3. It was developed at the International Center for Aggregate Research (ICAR), at The University of Texas at Austin. The instrument is appropriate for moderately and highly flowable fresh concrete with a slump value greater than 50–75 mm, especially SCC. It tests the rheological properties of fresh concrete and mortar with a maximum aggregate size of 40 mm, depending on the radius of the out cylinder. The ICR rheometer is designed based on a wide-gap, coaxial cylinder rheometer. It is made of a driver head that contains an electric motor and torque meter, a cross vane that is held by the clamp on the driver acting as an inner cylinder to prevent slip, a frame to affix the driver-vane assembly to the top of the cylinder, a laptop to operate the driver, record the torque during the test, and compute the rheological parameters, and a fresh concrete container which sticks many vertical strips around the inside wall to prevent slipping of the fluid along the container perimeter during the test. The size of the cylinder and the length of the vane shaft are chosen according to the nominal maximum size of the aggregate. The vane height and diameter are both 127 mm. ICAR rheometer can perform two types of tests, i.e., stress growth test and flow curve test. During the stress growth test, the cross vane is rotated, for example, at a low velocity of 0.025 rev/s. The initial torque increase is measured in terms of time. The maximum torque measured in this test is used to determine the static yield stress. The flow curve test calculates the dynamic yield stress and plastic viscosity. At the beginning of the trial, the vane is rotated at the maximum velocity to break down the possible thixotropic structure and ensure a consistent shearing history for each test sample. Then the vane speed slows down in a specified number of steps. The vane speed is kept constant for each stage, and both the average speed and the torque are logged. Usually, at least six steps are suggested. However, the number of steps can be chosen by the experimenter. Afterward, the plot of torque versus rotation speed of the vane is drawn. The rheological model parameters are calculated based on the least square regression of its transformation equation. 6.2.1.2 Couette rheometer One typical Couette rheometer is ConTec Viscometer 5 (see Figure 6.4). It is one of the highly improved models of the BML rheometer. The outer cylinder (container) rotates at a series of given angular velocities during the measurement, while the inner cylinder remains

Concrete rheometers  127

Figure 6.4  ConTec Viscometer 5.

Figure 6.5 Schematic cross section of the ConTec Viscometer 5 (Heirman et al., 2009a).

still. The inner cylinder is composed of two parts, as shown in Figure 6.5. The upper unit is stationary and records the applied torque from the test material, and the lower unit is fixed guard cylinders to eliminate the influence of the end effect (which will be discussed in Section 6.2.1.4) from the measurements. This specially designed geometry ensures that the material in the gap between the upper part of the inner cylinder and the outer cylinder is subjected to the perfect Couette flow. During the test, the material is pre-sheared at the maximum rotation speed applied for 30 s, followed by a decrease in rotational speed in given steps from the highest testing speed to the lowest speed. The average of the torque and rotational speed measured for each step will be calculated, and one set of data points are generated, provided that the torque reaches a steady state.

128  Rheology of Fresh Cement-Based Materials

6.2.1.3 Principle One of the biggest challenges in coaxial cylinder rheometry is the Couette inverse problem, which is about how to convert the rheological measurement (T, N) to the fundamental rheological unit. So far, there are two approaches to this problem: one is the numeric method, and the other is an analytical method. Because most inverse problems have no analytical solution, the numeric method is usually the only approach. Although this method does not need to assume the exact rheological expression for fluid, excellent programming skills and extensive theoretical knowledge about the inverse problem are required, which are challenging. On the other hand, by assuming the flow distribution of fluid and the geometry of the rheometer, the analytical method attempts to derive an equation between the rotational speed and torque for some simple rheological models. The transformation equations then can fit the test data to get the model parameters. Generally, the fitting result can be highly accurate when using an appropriate model. Therefore, the analytical method is the first choice for the Couette inverse problem. The actual flow of fresh concrete in a coaxial cylinder rheometer is very complicated, so some presumed condition is necessary to simplify the problem:

a. The material is considered a homogeneous fluid and is invariable with time during the test. b. Only the material in the space between the two cylinders (annulus) is considered. c. The material in the annulus is in a stable and laminar flow state, independent of the vertical direction, and the test is executed under equilibrium conditions. d. The flow is purely circular, and there is no flow in the radial direction. e. Any end effects at the top or bottom of the cylinder are ignored. f. The inertial effects are ignored. g. There is no slippage between a cylinder and the material. The velocity of the adjacent material is equal to the one of the cylinder.

The top view of a coaxial cylinder rheometer is shown in Figure 6.6. For the Searle rheometer, the shear rate of a rotation flow at radius r in cylindrical coordinate is (Macosko, 1994):

γ = r

dω (6.1) dr

According to the Cauchy stress principle (Malvern, 1969), the shear stress at any distance r in the annulus with the height of h can be expressed as follows:

τ=

T (6.2) 2πr 2 h

Equation (6.2) indicates that the shear stress gradually decreases as the distance from the axis of the rheometer increases. Because fresh concrete is a yield stress material, the material will not flow until the shear stress applied to the material exceeds the yield stress. As the shear stress grows, the adjacent material near the inner cylinder starts to flow. Under this condition, only part of the material close to the inner cylinder (shear zone) undergoes the shear flow. The shear flow stops at a certain imaginary radius r = Rp, where the shear stress is equal to the yield stress. The material outside this imaginary radius remains still (plug zone). As the shear stress keeps on growing, the region where the shear flow occurs

Concrete rheometers  129

Figure 6.6 Top view of a coaxial cylinder rheometer and flow in the annulus: (a) shearing at low rotating speed and (b) shearing with high rotation speed.

is expanded outward. The boundary between the sheared zone and the unsheared zone is called the plug radius Rp. If the shear stress is large enough, the shear zone will reach or even theoretically become larger than the outer cylinder. Here, the material in the entire annulus is sheared. The material in contact with the surface of the inner cylinder has the same angular velocity as the inner cylinder. Based on Eq. (6.1) and Eq. (6.2), the constitutive equation of the Bingham model can be written as follows:

T  dω  = τ0 + η  r (6.3) 2  dr  2πr h

Rearranging the above equation leads to Eq. (6.4):

 T 1 − τ 0  dr = −ηdω (6.4)  2 r 2πr h

Suppose all the material in the gap flows, the inner cylinder rotates at a constant angular velocity of Ω, and the out cylinder remains stationary. Equation (6.4) is integrated across the entire flow zone (the whole annulus), from ω = Ω at r = Ri to ω = 0 at r = R s:



Ro

Ri

τ   T − 0  dr = −  2πr 3h r 





0

η dω (6.5)

The result of the above integration is:

Ω=

T  1 1  τ R  − 2  − 0 ln  o  (6.6)  2 4πhη  Ri Ro  η  Ri 

130  Rheology of Fresh Cement-Based Materials

Equation (6.6) is the transformation equation of the Bingham model, named as Reiner– Riwlin equation. It is customary to replace the angular velocity Ω (in rad/s) with the rotational velocity N (in rps), Ω = 2πN.

N=

1  τ T  1 R  − − 0 ln o (6.7) 8π 2 hη  Ri2 Ro2  2πη  Ri 

or



R  4πh ln  o   Ri  8π h T= ηN + τ 0 (6.8)  1  1 1  1   R2 − R2   R2 − R2  i o i o 2

Although Eqs. (6.7) and (6.8) are derived from the Searle-type rheometer, they can be used for the Couette rheometers. If the material in the entire annulus is in a shear flow state, the relationship between rotational velocity (N) and torque (T) can be defined as N = AT − B. Therefore, the Reiner–Riwlin equation can be plotted as a straight line with its slope defined in terms of plastic viscosity, cylinder radii, and cylinder height, and the intercept defined in terms of yield stress, viscosity, and cylinder radii. Given the rheometer data versus rotational velocity, the slop (A) and intercept (B) of the straight-line fit can be determined and transformed into yield stress (τ0) and plastic viscosity (η) using the following equation:

η=

1  1 1  − 2  2  8π hA Ri Ro 

τ0 =

2

2πη B R  ln  o   Ri 

(6.9)

The same approach can apply to the Herschel–Bulkley fluid and the modified Bingham model. Heirman et al. (2006) gave a transformation equation for the Herschel–Bulkley fluid, which is written as follows: 1/ n   T τ0   1  T −   n − φˆ  1 − ,1,    2 2  n  2πRi hτ 0  2πRi hK K  



τ   T − − 0  2πRo2 hK K 

1/ n

  1   T ,1,    = 4πN (6.10)  n − φˆ  1 − 2  n    2πRo hτ 0 

However, the derivation was incomplete in literature (Heirman et al., 2006), due to that no good algorithm can be used to obtain the rheological parameters using Eq. (6.10). Heirman gave an approximate solution for the Herschel–Bulkley model:



R  4πh ln  o   Ri  22n+1 π n+1hK T= τ0 + N n (6.11) n  1 1    1 1 nn  2/ n − 2/ n   R2 − R2  i o  Ri Ro 

Concrete rheometers  131

Recently, Liu et al. (2020) derived the transformation equation of the Herschel–Bulkley model in case the material in the gap is entirely sheared, which can be presented as follows:

1 1    1 1  1 1 n n T  1 2πhτ 0 Ri 2   1 2πhτ 0 Ro2   T  ⋅ 2 F1  − , − ;1 − ; ⋅ 2 F1  − , − ;1 − ; − n    2 2    2πhKRi  n T n T  n n   2πhKRo   n n    N=  4 π (6.12)

Feys et al. (2013) proposed a transformation equation for the modified Bingham model, which is shown in Eq. (6.13):



R  4πh ln  o   Ri  ( Ro + Ri ) cN 2 (6.13) 8π 2 h 8π 3 h T= τ0 + µN + 1  1  1  ( Ro − Ri )  1  1  1  R2 − R2   R2 − R2   R2 − R2  i o i o i o

Li et al. (2019) suggested that Eq. (6.13) is only an approximate solution and derived the transformation equation in case the entire material in the gap flows under shearing. But generally, these equations are too complex to be applied in practice; thus, the Reiner–Riwlin equation is still the first choice for fresh concrete. 6.2.1.4 Measuring errors and artifacts Interpreting data from the rheological test is quite complicated. This section discusses three major sources of measurement or interpretive errors, namely, plug flow, particle migration, and end effect. 1. Plug flow If only a part of the material is sheared, the shear stress at the boundary Rp equals the yield stress τ0. The velocities at the boundary Ri and Rp are Ω and 0, respectively. For the Bingham fluid, integrating Eq. (6.4) over the flow domain is:



Rp

Ri

τ   T − 0  dr = −  2π r 3h r 





0

η dω (6.14)

The result of the above integration turns out to be:

Ω=

1  τ T  1  Rp  − 2  − 0 ln  (6.15) 2  4πhη  Ri Rp  η  Ri 

Replacing the τ in Eq. (6.2) with τ0, and the angular velocity Ω with the rotational speed N (N = Ω/2π, in rps), Eq. (6.15) can be written as follows:

N=

  τ T T  1 2πhτ 0  − − 0 ln  (6.16)   2 2 2 T  4πη  2πhτ 0 Ri  8π hη  Ri

132  Rheology of Fresh Cement-Based Materials

Figure 6.7  The rotation speed versus torque curve for a typical Bingham fluid.

From the above discussion, it can be seen that the N–T function of the Bingham model in coaxial rheometer can be plotted as a stepwise curve (see Figure 6.7), with a straight line with an initial curved part. The range of the nonlinear function depends on the dimension of the gap, yield stress, and height of the cylinder. The parameters of the Bingham model can be estimated by applying piecewise fitting in different sections of the test data. However, it is not easy for most engineers. Considering Eq. (6.7) is valid only if the shear flow occurs to all the material in the gap, it means that the plug radius should be at least equal to the radius of the outer cylinder. According to Eq. (6.2), the minimum torque for this will be T = 2πRo2 hτ 0. Therefore, a straightforward method to obtain the Bingham parameters is by using Eq. (6.7) to all the test data, calculating the minimum torque for the entire shear flow, eliminating the invalid data, and repeating the above procedures until all the data left are qualified for data fitting. Wallevik (2001) and Heirman et al. (2009b) proved that neglecting the plug flow in the ConTec Viscometers induces slight error for Bingham and Herschel–Bulkley materials, respectively. 2. Particle migration Although fresh concrete is considered as a homogeneous material, the density difference between coarse particles and mortar matrix exists. During the rheological test, coarse aggregates are pushed away from the zone near the inner cylinder because of the highest momentum (shear rate) (Wallevik, 2003), and particle migration may occur. The severity of the particle migration decreases with the shortened test duration time, low shear rate, relatively narrow gap size, low yield stress, and high plastic viscosity of the material. Particle migration can cause the layer near the inner cylinder to get devoid of coarse aggregate, and then the material becomes heterogeneous. Especially for large gap sizes and high yield stress, particle movement from the sheared zone may increase the range and packing density of the plug zone (Feys and Khayat, 2013). Wallevik et al. (2015) proposed a simple method to evaluate the influence of particle migration. The procedure involves the calculation of the thickness of the sheared zone via plug flow identification. If the width of the shear zone is smaller than or near the maximum size of the aggregate for the majority of time during the rheological measurement, it may imply a remarkable particle migration. Here, significantly lower rheological properties can be observed, and the rheological measurements would no longer be valid.

Concrete rheometers  133

3. End effects The derivation of transformation equations and related discussions presented above were based on the distribution of shear rate and shear stress in the gap of the coaxial cylinder rheometers. However, the inner cylinder should be immersed in the testing material. Because the material is filled above and below the inner cylinder and sheared during the testing, extra shear stress will be added to the total recorded torque. These end effects must be eliminated when calculating the rheological parameters. One approach is to change the geometry of the rheometer so that shear stress applied to the cylinder ends can be minimized or eliminated. For Couette rheometers like ConTec 5 or BML Viscometer, fixed guard cylinders can be placed below (and above, if necessary). These guard cylinders are adjacent to the inner cylinder but not connected to it, allowing the inner cylinder to rotate at minor angles while measuring the stress exerting on the side. However, this solution is not suitable for a Searle-type rheometer. A double-gap coaxial cylinder rheometer is an option since the cylindrical surface is much larger on the sides than on its ends. Thus, the end effects can be omitted. Yan and James (1997) suggested that the volume below the cylinder can be considered as a parallel-plate rheometer while the volume in the annulus as a coaxial cylinder rheometer. Therefore, the total torque equals the torque from the coaxial cylinder rheometer part, plus the torque from the parallel-plate rheometer part. This method applies to the FHPCM rheometer for fresh concrete. In order to determine the contribution of end effects to the total torque, Dzuy and Boger (1985) compared three approximate solutions with experimental measurements. They used three different distribution assumptions of shear stress acting on the end of the vane. One is the uniform distribution, the other is the power-law distribution, and the third is the prior unknown distribution. They determined that the third one is the most accurate method. However, this method needs at least two different vanes to make an experimental measurement. Whorlow (1992) suggested measuring the torque at a given rotational speed with the gap of the rheometer filled to a different height. Because the end effects are the same for all the tests, the plot of torque versus immersed vane height should be a straight line. The intercept represents the amount of torque caused by the end effect. This method is also suitable for fully immersed cylinders of different heights. Nevertheless, end effects usually vary for different materials. Individual calibrations are needed for almost every material with at least two vanes, making the test procedure very complicated. Given this problem, Laskar et al. (2007; 2011) established a new transformation equation, which takes the end effects into account. Figure 6.8a and b illustrates the shear rate distribution profile with a stationary surface in horizontal and vertical directions. The geometrical dimensions of the vane and cylinder are shown in Figure 6.8c. The total torque exerted on the inner cylinder can be considered as the summation of the torques of all regions. 3.1 Volume CDBA (V1) Supposing the angular velocity of the vane is ω, the shear (strain) rate along the radial direction is

γ = ω Ri /Ri = ω (6.17)

The torque contribution is

T1 = (τ 0 + ηω ) 2πRi2 H (6.18)

134  Rheology of Fresh Cement-Based Materials

Figure 6.8 (a–c) Velocity profile along with the horizontal and vertical directions (Laskar and Bhattacharjee, 2011).

3.2 Volume DPOB (V2) Assuming there is a circular element dr along BD at radius r from the rotational axis of the vane, the linear velocity at this radius is equal to rw, and the shear rate is γ = ω r Z2 , the torque on this elemental disc can be written as follows:

dT = (τ 0 + µγ ) 2πr 2 dr (6.19)

The torque T2 is:

T2 =



Ri

0

ωr  2πRi3 πRi4ω  2  τ 0 + η Z  ⋅ 2πr dr = 3 τ 0 + 2Z η (6.20) 2 2

3.3 Volume KCAL (V3) Similar to 3.2, the torque T3 can be expressed as follows:

T3 =



Ri

0

ωr  2πRi3 πRi4ω  2 η (6.21) ⋅ π = τ + η 2 τ + r dr 0 0  Z1  3 2Z1

3.4 Volume of hollow-cylinder IHPD-BOGN (V5) As shown in Figure 6.9, for an elemental layer of thickness dz at a height z from the bottom of the cylindrical surface DPOB, the velocity along the radial direcvz z ω Ri = . Define the effective tion on the surface of DPOB is written as: vr = Z2 Z2

Concrete rheometers  135

Figure 6.9 Velocity profile in the gap between vane and outer cylinder (Laskar and Bhattacharjee, 2011).

gap of annulus g as a constant of Ro−Ri; the shear stress at the height of z from vz . The force on the elemental area will be the bottom is given as τ r = τ o + µ Z2 g  vz  dF =  τ 0 + µ  ⋅ 2πRi dz , and the total force is given as follows: Z  2g 

F=



Z2

0

 ηv  dF = 2πRi  τ o + Z2 (6.22) 2 g  

And the torque of the part is:



 ω Riη  T4 = Ri F = 2πRi2 Z2  τ 0 + (6.23) 2 g  

3.5 Volume of hollow-cylinder FJCK-LAME (V4) The torque T4 can be deduced in a similar approach as in 3.4 above, and the result is shown in Eq. (6.24):

T5 = R1 ⋅



Z1

0

 ω Riη  dF = 2πRi2 Z1  τ 0 + (6.24) 2 g  

3.6 Volume in the gap of JIDC-ABNM (V6) The velocity and shear rate on the surface of CDBA are v = ωRi and v/g, respectively. The torque of this part is:

 υ T6 =  τ 0 + η  ⋅ 2πRi2 H (6.25) g 

Assuming that the vane is at the center of the testing material, Z1 = Z 2 = Z, and T2 = T3, T4 = T5. According to the literature (Barnes and Carnali, 1990, Dzuy and Boger, 1985, Sherwood and Meeten, 1991), material in volume V1 does not shear during the test. Therefore, T1 = 0. The summation of the above torques of different parts results in Eq. (6.26):

T = 2T2 + 2T4 + T6 (6.26)

136  Rheology of Fresh Cement-Based Materials

Figure 6.10  Serrated shapes of the inner cylinder of the ConTec Viscometer 5.

By replacing T2 , T4, and T6 in Eq. (6.26) with Eq. (6.20), Eq. (6.24), and Eq. (6.25), the following relationship can be derived:

R π 2 Ri3  R1 H+Z  H T = 4πRi2  + Z + i  τ 0 + + η N (6.27)  2 3 15  2Z Ro − Ri 

The equation above shows that the total torque T is a linear function of rotational speed N. Therefore, the Bingham parameters can be calculated by fitting the experimental data with Eq. (6.27). 4. Hydrodynamic pressure In order to avoid slippage between the concrete and the rheometer, the internal wall of the rheometer needs to be designed with serrated shapes (Figure 6.10), allowing the coarse aggregate to be part of the internal boundary. The coarse aggregates will sit in the space between the serration, and no slippage boundary can be observed. However, well-fined flow occurs between the blades for the four blades-vane rheometers, introducing extra errors in calculating the shear rate. Wallevik (2014) indicated that hydrodynamic pressure would apply to the blades’ wall boundary for a four blades-vane rheometer and contribute to torque, just like the viscous stresses. Therefore, the total torque is the summation of the torque caused by viscous stress and the torque caused by hydrodynamic pressure. Numerical simulation suggested that hydrodynamic pressure

Concrete rheometers  137

accounts for 80% of the total torque recorded by the rheometer. The conclusion is valid for the Newtonian, Bingham, and Herschel–Bulkley fluids. This emphasizes the necessity of thoroughly studying the influence of hydrodynamic pressure on different types of rheometers.

6.2.2 Parallel-plate rheometer A parallel-plate rheometer is generally used to measure the rheological properties of polymeric materials. The distance between the two plates is usually less than 1 mm. However, it is not applicable to concrete because of particles, sand, and coarse aggregates. The distance between the two plates should be 5–10 times the maximum particle size of concrete mixture, i.e., 50–100 mm. One solution is to fill the concrete in a cylindrical container with a blade rotation. A seal is needed to keep the material between the blades to ensure the material is sheared without a leakage. Because the container introduces extra shear resistance to the material, the analytical solution to decide the shear rate and shear stress cannot be applied. One typical parallel-plate rheometer for concrete is BTRHEOM, developed at LCPC (Laboratoire Central des Ponts et Chaussées). The rheometer is specially designed to measure moderately to highly flowable concrete (slump at least 100 mm, up to SCC). By assuming complete concrete slippage at the wall of the container, its influence can be ignored. The maximum size of aggregate is up to 25 mm. It holds about 7 L specimens of concrete and can be used in the lab and on a construction site. 6.2.2.1 Geometry The prototype of the BTRHEOM rheometer is illustrated in Figure 6.11. It consists of a 120 mm radius hollow cylinder with two parallel blades installed at the top and bottom of the cylinder. The vertical distance between the blades is 100 mm. The bottom blade is fixed as the top blade rotates. Both the blades are designed with openings, so that slippage can be avoided. A motor is mounted below the cylinder and connected to the top plate vertically with a shaft. The radius of the shaft is 20 mm. A vibrator can apply to the sample material in the container to either consolidate concrete or measure the vibration’s effect on the rheological parameters. However, the rheological measurement cannot be carried out during the vibration. As the top blade rotates at a series of different rotational velocities, torque caused by the resistance of the concrete being sheared is measured through the top blade.

Figure 6.11 (a–c) Prototype of the BTRHEOM rheometer (Hu et al., 1996).

138  Rheology of Fresh Cement-Based Materials

An accompanying software program, ADRHEO, operates the rheometer (rotation speed and vibration), collects the measurements (torque and rotation speed), and calculates the rheological parameters from the raw data. The maximum measurable torque value is about 14 N·m. The range of rotation speed is from 0.63 rad/s (0.1 rev/s) up to 6.3 rad/s (1 rev/s), though 0.63 rad/s (0.1 rev/s) and 5.02 rad/s (0.8 rev/s) are chosen as the lower and upper limits for testing, respectively. A text output file will be generated after the test, including the torque values at each rotation speed and the calculated rheological parameters, either the Bingham or the Herschel–Bulkley parameters (depending on the different versions of ADRHEO software). 6.2.2.2 Principle Before initial testing, the BTRHEOM rheometer needs to be calibrated for rotational speed, torque, and vibration frequency according to the operation manual. A seal is used to ensure that no concrete flows into the region between the base container and upper rotating cylinder before each testing. Then a rotational calibration test is executed for further refinement. For each set of seals, a rheology test is performed with water to determine the frictional resistance of the seal. The ADRHEO software uses the results to eliminate the friction effects of the seals in the following rheological test of concrete or mortar samples. The ADRHEO software controls the entire testing process. After the container is filled with the testing material, an optional operation is to vibrate it for 15 s to consolidate the concrete (vibration can also be applied during the measurement). The frequency of this previbration ranges from 35 to 55 Hz. It should be noted that pre-vibration is not for concrete with low yield stress, such as SCC. Then measurement starts. The test is composed of one or two sequential down ramps (up ramps are also possible but rarely used, except for the thixotropy test). Each down ramp includes five to ten measurement points, i.e., torques at decreasing rotation speed are collected. For each measurement data point, the rotation speed remains constant for about 20 s so that a torque measurement can be stabilized and recorded. Under the no-slip boundary condition, the rotational velocity at a radius r is Ωr on the upper blade and zero on the lower blade in cylindrical coordinates. The shear rate at radius r can be written as follows:

γ (r) =

Ωr (6.28) H

Equation (6.28) indicates that the shear rate alters along the radial direction, and simple shear can be achieved at any radius r in a parallel-plate rheometer. As a result, the local shear rate in between the two blades is known no matter what the material constitutive law is. It can also be shown that the shear stress τ(R) at the edge of the geometry (r = R) can be presented as follows:

τ (R) =

T 2πR2

Ω ∂T   3 + T ∂Ω  (6.29)

ΩR . Because of the derivH ative term of the torque regarding the rotational velocity in Eq. (6.29), numerous accurate T(Ω) data need to be collected to calculate τ = (γ ), which is relatively difficult to use. Let r = R, and the share rate at the edge of the geometry is γ (R) =

Concrete rheometers  139

Another approach is to replace the shear stress and shear rate in a rheological model by Eqs. (6.28) and (6.29), and establish a transformation equation between the torque T and the rotational speed N. The rheological properties can be estimated via data regression from a single T(Ω) measurement. According to the experiment, the relationship between the torque T and the rotation speed N can be expressed as follows:

T = T0 + AN n (6.30)

where T0 is the flow resistance, A is the viscosity factor, and n is the flow index factor. This expression is in a similar function form to the Herschel–Bulkley model. Suppose fresh concrete is a Herschel–Bulkley material; its constitutive equation is τ = τ 0 + Kγ n . If the power exponent index equals 1, the Herschel–Bulkley model will be turned into the Bingham model (τ = τ 0 + ηγ ). Integrating the contribution of all the surface elements to the torque leads to the following equations (De Larrard et al., 1998): 2π 3  3 T0 = 3 R2 − R1 τ 0  (6.31)  (2π)n+1  n +3 n +3  A = (b + 3)hn R2 − R1 K

(



)

(

)

The above equations can be inverted to get the material parameters by nonlinear regression of the T–N curve, which results in:



3  τ 0 = 2π R3 − R3 T0 2 1  (6.32)  hn (n + 3)  K = 0.9 (2π)n+1 Rn+3 − Rn+3 A 2 1 

(

)

(

)

For the Bingham fluid, the yield stress would be τ0, and the equivalent plastic viscosity η would be:

η=

3K n−1 γ max (6.33) n+2

Ωmax R2 is the maximum shear rate recorded in the measurement. h One thing that should be mentioned here is that the BTRHEOM cannot shear the specimen adequately at the angular velocity of 0.015 rad/s (0.1 rev/s) or lower. The data point will be excluded as an outlier for data analysis.

where γ max =

6.2.2.3 Measuring errors and artifacts One advantage of using the rheometer is that the entire material on a horizontal plane is sheared uniformly for a given loading condition. As a result, plug flow, which often appears in coaxial cylinder rheometers, does not occur in the BTRHEOM rheometer.

140  Rheology of Fresh Cement-Based Materials

The shear-induced sedimentation of particles is a possible measurement artifact. Note that the vertical gap between the two parallel plates is about 100 mm, and then the gravity effect caused by the difference between the cement–sand mix and aggregate cannot be ignored. During the rheological test, particles, especially coarse aggregates, tend to settle down from the zone near the upper plate. Other possible artifacts include the slip risk and wall effect. Due to the limited volume of sheared specimens, these artifacts are minimal. Therefore, no considerable systematic error will be introduced after careful calibration. Considering all the friction-related effects, a 10% correction applies to factor K, as it is usually done for plastic viscosity (Hu et al., 1996). Another interesting point is about the calculation of Bingham parameters. Although Eqs. (6.32) and (6.33) can be used to obtain the yield stress and plastic viscosity, these two rheological parameters can be calculated directly from Eq. (6.30) provided n = 1. Unfortunately, the calculation results are different using these two approaches. Therefore, further research is needed to decide which method is better.

6.2.3 Other rheometers 6.2.3.1 CEMAGREF-IMG rheometer The CEMAGREF-IMG rheometer is a large coaxial cylinder rheometer, and it can load about 500 L of concrete for testing (see Figure 6.12). During the test, the inner cylinder rotates, while the outer cylinder remains stationary. The surface of the inner cylinder is made of a metallic grid to minimize the slippage of concrete, and vertical blades are welded on the internal wall of the outer cylinder. A rubber gasket is attached to the bottom of the inner cylinder to prevent material leakage from the slit between the bottom of the container and the cylinder. However, because of its large volume, some plug flow is expected when testing concrete materials. Therefore, Eq. (6.7) can calculate the rheological parameters of the Bingham material. Alternatively, Eq. (6.16) should be used if plug flow occurs. 6.2.3.2 Viskomat XL Schleibinger developed Viskmat XL based on many years of experience with rheometer for mortar and fresh concrete. The Viskmat XL is a vane rheometer, which can rotate in both clockwise and counterclockwise directions (See Figure 6.13). Unlike the ICAR rheometer, the Viskomat XL has frame vanes. Usually, it is run under a shear rate-controlled mode, and the rotational speed can be programmed in linear steps, either increasing or decreasing. Also, a logarithmic or oscillating mode is possible as an option. Alternatively, it can be run in a shear stress-controlled mode. Therefore, the torque can be predefined over time, so the rotational speed can automatically be controlled to achieve the preset torque (shear stress). Viskomat XL equips a double-wall container, where cooling liquid can circulate between the walls. Via this, the temperature of the specimen can be controlled. It also has a high time resolution, and the sample rate can be set from 0.005 s to 10 min. Overall, it is a versatile apparatus for the rheological measurement of mortar and fresh concrete. 6.2.3.3 The IBB rheometer The IBB rheometer is based on the existing device (MKIII) developed by Tattersall (see Figure 6.2b). It is completely automated and uses a computer program to drive an impeller to rotate in fresh concrete. The software can analyze the testing results and calculate the rheological parameters of the Bingham material.

Concrete rheometers  141

Figure 6.12  The CEMAGREF-IMG rheometer (Banfill et al., 2001).

Figure 6.13 The Viskmat XL.

142  Rheology of Fresh Cement-Based Materials

Figure 6.14  Rheometer developed by Yuan and Shi.

6.2.3.4 Rheometer developed in China Yuan and Shi have developed a prototype rheological instrument, which is basically a vane rheometer (Yuan and Shi, 2018). It is commercialized by a Chinese company a few years ago (see Figure 6.14). In comparison with other rheometers, this one is more automated. After loading concrete in the container, the shaft will drop automatically to a specified height and start rotation in rampway to build a torque–rotation speed curve. The fresh concrete specimen is first sheared in the highest speed to eliminate thixotropy, and then ramp downwards. The Bingham model and other rheological models can be used to transform torque–rotation speed data into yield stress and viscosity by software embedded in the rheometer. A cylindrical rotator is also equipped to measure the properties of the lubrication layer. And the software has a friendly interface for each function, which can facilitate the operation of the rheometer. 6.2.3.5 The modifications of the BTRHEOM rheometer As for the parallel-plate rheometer, the UIUC concrete rheometer was built at the University of Illinois (Beaupré et al., 2004). The significant improvements are to help install and clean

Concrete rheometers  143

the device during the test. Besides, other researchers modified the BTRHEOM (Struble et al., 2001, Szecsy, 1997), but these modified rheometers have not been commercialized. The major improvements include suspending the top plate from a shaft held above the concrete and eliminating the center axis inside the concrete container. 6.2.3.6 Other instruments Laskar and Bhattacharjee (2011) of India developed a cross vane rheometer; however, more researchers focused on designing the spindle with a different geometry. For example, Gerland et al. (2019) modified Viskomat NT with a ball probe. They used a simulationbased approach to determine the yield stress and plastic viscosity of UHPC and SCC from the rheometer measurement. Soualhi et al. (2017) from France also designed a portable rheometer with new vane geometry. Other spindle types include helical, H-shape, and double spiral. Although rotational rheometers can give an accurate description of the rheological behavior of fresh concrete, they are not suitable for online and continuous applications. On the other hand, the rheological measurements of fresh concrete during mixing and transportation have aroused wide concern. One example is the attempt to correlate the output of the concrete-mixing truck to values obtained by rotational rheometers. The output could be a watt meter or hydraulic pressure, which is known as the “slump meter”. However, the experimental errors might be higher than usual, making it more difficult to characterize the rheological values of the concrete-mixing truck. Wallevik et al. (2020) used a series of computer simulations to analyze the relationship between the power required to rotate the drum of a concrete-mixing truck and the rheological properties of fresh concrete. They found that the power could be calculated using different Bingham parameters (i.e., yield stress τ0 and plastic viscosity μ). With those simulations, the rotation speed–power of the drum curves were plotted, and the intersection value G and the slope H were calculated, which can be used to represent the truck’s rheological properties. By power used per unit mass, a possible relationship between the truck’s rheological values G and H and the Bingham parameters τ0 and μ could be established. Although these two equations are not completely accurate, it proves that using the vehicle as a rheometer is technically feasible.

6.3 MEASURING PROCEDURES

6.3.1 Preparation of specimen For rheological testing, the mixing process of concrete materials is similar to those for other properties. The concrete specimen should be homogeneous so that stable and reproducible rheological measurements can be achieved during the test. In addition, adequate raw materials should be prepared to ensure the test accuracy; usually, the capacity of the mixture should be at least two times the volume of the container for rheological testing.

6.3.2 The testing procedures of ICAR The operation interface of ICAR rheometer software is shown in Figure 6.15. All the operations can be managed through a single screen. In this interface, you can specify the name of the storage file and its storage path, set the geometry parameters, and execute the stress growth test and flow curve test. If something goes wrong, the software can terminate the trial.

144  Rheology of Fresh Cement-Based Materials

Figure 6.15 The ICAR rheometer software interface.

The testing procedures for stress growth test include the following steps: 1. Assemble all the components of the ICAR rheometer correctly according to the manual; 2. Place the testing mixture into the container. Ensure the mix and the vertical ribs on the internal wall of the container have the same height. Insert the cross vane into the mix; 3. Ensure the default geometrical parameters set in the operation interface are the actual sizes of the vane and the container; 4. Input the rotation speed value. The vane speed for the stress growth test is typically between 0.01 and 0.05 rps, with 0.025 rps by default; 5. Click the “Reset” button to ensure the initial torque of the vane shaft is zero; 6. Start the test; 7. Once the peak of the torque–time curve appears and the torque gradually decreases, click the “Finish” button to complete the test. The testing procedures for the flow curve test are as follows:

1. After the stress growth test is completed, the flow curve test of the mixture can be carried out; 2. In order to destroy the flocculation structure of the mixture and provide a consistent shearing history, it is necessary to let the vane rotate at maximum speed. Usually, the breakdown time and the breakdown speed are set to 20 s and 0.50 rps, respectively, before starting the flow curve test; 3. Set the initial rotational speed for the flow curve test as 0.50 rps and the end speed as 0.05 rps, which is divided into seven test points evenly, each of which lasts for 5 s; 4. Click the “Start” button to start the flow curve test; 5. If the mixture is so viscous or the yield stress is so high that the vane cannot rotate at the maximum speed of 0.50 rps, immediately click the “Abort” button to stop the test;

Concrete rheometers  145

Figure 6.16 The assembly of the inner cylinder of Viscometer 5 (Banfill et al., 2001).

Figure 6.17  The main operation interface of ConTec Viscometer 5.

6. Once the test is finished, the flow curve will be plotted and illustrated in the interface windows, and the linear fitting and Bingham parameters will be calculated. Then, testing data will be written into an output file for further analysis.

6.3.3 The testing procedures of ConTec Viscometer 5 Before testing, the inner cylinder should be mounted to the axis of the testing unit, as illustrated in Figure 6.16. Each test takes about three minutes, and the whole test is controlled by a special software named FRESHWIN (see Figure 6.17). During the trial, the specimen is sheared for about 1 min. Then, the container can be emptied for the next test. All the parameters needed for testing can be inputted into the software as a basic setup. As the test is finished, the output of the test result will be plotted (see Figure 6.17), and the rheological parameters in the fundamental unit will be calculated.

146  Rheology of Fresh Cement-Based Materials

Similar to the flow curve test of the ICAR rheometer, the flow curve test uses seven points to calculate the rheological parameter. The rotational speed equally decreases between the highest and lowest speeds. For each point, the testing period is 5 s, including the first 1.5 s transient interval and 3.5 sampling interval. Then, rotational speed increases to the maximum speed for 2 s afterward and continues to rotate for 5 s at 2/3 maximum speed so that the aggregate segregation factor can be evaluated (segregation test). The testing procedures include the following steps:

1. Fill the concrete mixture into the container, and lower the inner cylinder until it inserts into the container to a fixed depth. Ensure the mix has the same height as the stripes welded on the internal wall of the container. 2. Execute the FRESHWIN software, select “Process” from the menu, and select “Parameters” from the drop-down menu. Choose the existed test name in the “Name” input box, or click the “Add” button to set up a new test. 3. Ensure to input the correct values in the “Cylinder dimensions” group box. By default, the height of the inner cylinder is 0.2 m, the radius of the inner cylinder is 0.1 m, and the radius of the outer cylinder is 0.145 m. 4. Select the correct equation in the “Equation” group box. “Reiner-Rivlin equation” can be applied in most cases. 5. Input proper parameters in the “Run time parameters” and “Beater Control” group boxes. 6. Click the “OK” button, close the process parameter setting, and return to the initial interface of the FRESHWIN software. Click the “Start” button to execute the test. The inner cylinder will descend and insert into the mix. After the pre-shearing, the flow curve test will be conducted and followed by the segregation test. 7. Select “Save” from the “File” menu, and store the data on the disk after finishing the test. 8. The inner cylinder will rise automatically, and the container will be taken off, emptied, and ready for the next test. If time permits, the inner cylinder should be disassembled and cleaned after every two tests. For each test, the rheological parameters will be calculated automatically. Besides, the diagram of rotation velocity vs. torque will be drawn, and linear regression will be used to establish an equation between rotation velocity and torque. A particular point on the line as the rotation velocity is 2/3 of the maximum velocity (see Figure 6.18), will be calculated. Based on this, the segregation factor will be calculated using the following equation:

Seg =

H − H′ ⋅ 100% (6.34) H

The aggregate segregation test examines if the aggregate is segregated from the concrete mix. If the segregation factor Seg is  0.5 Pa/s) for cast-in-situ concrete formwork due to its longer dimensions and shapes. This is because its high structuration rate impedes the free flow of the material. It is recommended to use SCC having a high flocculation rate in precast concrete components (Rahman et al., 2014). To control the structuration rate without greatly affecting the slump flow, Roussel and Cussigh (2007) proposed five main factors to adjust the mix proportions: • • • • •

The total amount of the powders in the mixture. The weight ratio of water/powders (w/p) ratio. The fineness of the powders. The amount of superplasticizer. The amount of viscosity agent (VA).

Rahman et al. (2014) added fly ash and limestone powder (LSP) in different percentages into the SCC mixture and concluded that increasing the FA and LSP content increased the thixotropy of the mix. The results can be attributed to the increase in the structuration rate due to the increase in the fineness of the powders. Ahari et al. (2015a) investigated the thixotropic behavior of SCC by incorporating FA, SF, metakaolin (MK), class C fly ash (FAC), class F fly ash (FAF), and granulated BFS. The results showed that the increase in thixotropic behavior with time was higher in the case of SF and MK as compared with reference mixes and other minerals. In addition, recycled coarse aggregates also influence the thixotropy of SCC. Bir (2018) noted an increase in the degree of thixotropy of SCC mixture by increasing the replacement of natural aggregates with recycled aggregates. This was attributed to the high internal friction associated with the recycled aggregates. 8.2.2.2 Shear-thinning or shear-thickening behavior Shear thinning is the decrease in the apparent viscosity with shear rate, while shear thickening is the increase of apparent viscosity by increasing the shear rate. Generally, shear thickening is considered a potential industrial problem in terms of concrete production and casting. The popular theories to explain shear-thickening behavior include order–disorder transition theory, grain inertia theory of large particles, and clustering theory of fine particles. For the order–disorder theory, the flow is easy if particles are ordered into layers, while the disordered structure consumes more energy to flow due to the jamming of particles, increasing the viscosity. In the case of grain inertia theory

200  Rheology of Fresh Cement-Based Materials

of large particles, the momentum is transferred between the suspended particles and the dominance of grain inertia, depending on the particle’s Reynolds number. For the clustering theory, shear thickening could occur at certain critical shear stress where hydrodynamic forces due to flow become larger than the interparticle repulsive forces. De Larrard et al. (1998) and Feys et al. (2008b) believed that SCC showed a shear-thickening phenomenon. Chia and Zhang (2004) studied the shear-thickening phenomenon in SCC with AEA. The results showed that SCC mixtures exhibited a transition from shear thickening to shear thinning at 8.3% of air content. By increasing the air content, the adsorption of the bubble on cement particles could increase, thus hindering the formation of the cluster. The addition of air bubbles also enhances the deformation capacity of the mixture and reduces the hydrodynamic forces. Therefore, the mixture tends to show shear-thinning behavior at high air content. Huang et al. (2018) observed the shear-thickening behavior of SCC with the addition of SP. They proposed three reasons for this behavior: (1) SP disperses the fine particles in the solution. According to the clustering theory, the shear-thickening phenomenon is attributed to the change of particles from dispersion to cluster. As a result, the addition of SP facilitates the occurrence of shear thickening. (2) The increase of shear rate results in the disorder degree of fine particles and polymer chains, which leads to shear thickening following the order–disorder theory. (3) The adsorbed SP can be torn off from fine particles due to the shear stress, and the desorption state is more likely to form a cluster. Güneyisi et al (2016b) used tire chips and crumb rubber to prepare concrete, and they found that the exponent ‘n’ (the Herschel–Bulkley) values and ‘c/µ’ (the modified Bingham) coefficients increased by increasing the replacement level of natural aggregates with crumb rubber and tire chips, indicating the shear-thickening behavior. The highest ‘c/µ’ coefficients and exponent ‘n’ values were observed in the case of replacement of natural coarse aggregates with tire chips, and the lowest values were achieved when fine aggregates were replaced with crumb rubber. Shear-thickening behavior was observed when natural aggregates were used in SCC, but this behavior decreases when natural aggregates were replaced with lightweight aggregates (LWA) made with FA (Gesoglu et al., 2015). Due to the spherical shape of LWA, decreased yield stress and plastic viscosity were observed. Le et al. (2015) pointed out a decrease in the degree of shear thickening of SCC with the incorporation of RHA (MPS 5.7 µm) and SF. It was found that the effect of SF was much stronger than that of RHA. 8.3 FORMWORK PRESSURE OF SCC In comparison with conventional concrete, SCC has very high workability, which may result in large lateral pressure on formwork and cause economic and safety problems. If contractors and engineers design the formwork of SCC based on the full hydrostatic pressure without considering the formwork pressure variations during casting and pressure decay following placement, this will lead to increased costs for formwork systems—compromising profitability by offsetting the cost savings associated with SCC due to the rapid placement and labor savings. Form or molds receive SCC in different shapes and sizes. The fluidity and stability performance of the SCC mixture are influenced by the formwork characteristics. To avoid the honeycombing and surface defects, the formwork should be watertight and grout-tight, especially when the SCC has a low viscosity.

Rheology and self-compacting concrete  201

8.3.1 Factors affecting formwork pressure The lateral pressure exerted by concrete on formwork is influenced by the mixture composition, formwork characteristics, and placement conditions. Mixture composition includes the binder type and content, water-to-cementitious ratio, SCMs, fillers, paste volume, characteristics and content of coarse aggregates, chemical admixtures, concrete consistency, concrete unit weight, and temperature (Omran and Khayat, 2014). The most important factors that affect the formwork pressure are casting rate, concrete yield stress, and formwork height (Billberg, 2012, Geiker and Jacobsen, 2019). In the case of pumping concrete from the bottom, the full height of concrete will be in motion, and the resulting pressure will be the cumulative pressure of hydrostatic and pump pressure. If the concrete is not in motion, then the pressure can be reduced due to the thixotropic properties of SCC (Billberg, 2012). The phenomenon of thixotropy has an important influence on the various applications of SCC, such as formwork pressure development, pumpability, and multilayer casting (interface behavior of the successive layers). There could be a relationship between the lateral pressure of SCC and thixotropy. The greater the thixotropy, the lower the lateral pressure and the faster the rate of pressure drop with time. This is attributed to the faster re-gaining of the yield stress of the mixture when left at rest without any shearing action. The sharper drop of lateral pressure can be obtained by increasing the structural buildup. Omran and Khayat (2014) concluded that the greater the level of thixotropy, the greater the structural buildup at rest, and thus the lower the lateral pressure during placing and the more rapid the drop in pressure with time. The structuring rate of concrete has a considerable influence on the distribution of lateral pressure on the formwork, regardless of the time. A faster degree of structuring leads to the speedy development of cohesiveness soon after casting, and thus, a cohesive mixture exerts lower lateral pressure on formwork than its full hydrostatic pressure. At a faster flocculation rate (a fast structural buildup rate) and a lower casting rate, the maximum value of lateral stress remains lower than the hydrostatic pressure during casting (Ovarlez and Roussel, 2006). The structural buildup of cement-based materials has been extensively studied. It is found that mineral admixtures can increase the structure buildup rate by accelerating the hydration rate. However, a high hydration rate did not always lead to a high structural buildup rate (Yuan et al., 2017b). Kim et al. (2012) incorporated FA and limestone powder in SCC mixes and measured the relationship between flowability and lateral pressure on formwork. The results showed that the incorporation of both minerals increased the lateral pressure on formwork. The formwork pressure can be reduced by adding a small amount of processed clay, MK, and alumino-silica (Kim et al., 2010).

8.3.2 Formwork pressure prediction Due to the high casting rate of SCC, the formwork should be adequately designed to accommodate the expected liquid head formwork pressure. It is recommended to design the SCC formwork for a full liquid head, especially at a high casting rate. However, ACI 347R-14 includes new provisions and recommendations for casting SCC based on the rate of concrete placement relative to the rate of development of concrete stiffness/strength. The measure of stiffening characteristics of the SCC was included in the methods, and the methods are capable of being easily checked on-site using some easy measurements. The details of three typical methods for measuring the lateral pressure for SCC are described in the following.

202  Rheology of Fresh Cement-Based Materials

8.3.2.1 Method proposed by Gardner (Gardner et al., 2012) Based on some field testing results, Gardner et al. (2012) proposed a parameter—t 400 (time for the drop of slump flow to 400 mm)—to characterize the SCC. The time to reach zero slump flow (t 0) is defined as follows:

t0 = t400 ×

inital slump flow (8.1) initial slump flow − 400 mm

Using t 0, a simple equation was developed to estimate the development of lateral pressure (Pmax) with time. After the placement, the limiting value of lateral pressure Pmax (kPa) with time can be calculated as follows:

 t2  Pmax = wR  t − , 2t0  



Pmax = wRt0 2 ,

t < t0 (8.2)

t ≥  t0 (8.3)

where w is the unit weight of concrete (kN/m3), h is the height of placement (m), and R is the rate of placement (m/h). The equation cannot be used for the cases when the casting time is greater than that required to achieve Pmax. For t > t0, the pressure is assumed to remain constant at the maximum value. If the time to fill the form, th (= height of form/R), is less than t 0, t = th is used in the equation:

 t2  Pmax = wR  t − h  (8.4) 2t0  

It is worth mentioning that the experimental results in Gardner et al. (2012) are limited in that the maximum concrete head available to the authors was 4 m, implying the hydrostatic pressure of 96 kPa. 8.3.2.2 Method proposed by Khayat (Khayat and Omran, 2010) Khayat proposed several methods for measuring and predicting the formwork pressure. A portable measurement device, referred to as the UofS2 pressure column, was developed to evaluate the lateral pressure exerted by plastic concrete. The UofS2 pressure column is a polyvinyl chloride (PVC) cylindrical pressure vessel with flat-plate flange closures. During the test, the vessel is initially filled with concrete (from the top) to a height of 0.5 m above the centerline of the bottom sensor, at a given placement rate and without any vibration. The top of the pressure vessel is then closed and sealed, and the internal air pressure is gradually increased to simulate the hydrostatic head of the placement of additional SCC at the given placement rate. The corresponding lateral pressure exerted on the vessel wall was recorded using the sensor and plotted against the hydrostatic pressure. Another method is based on the structural buildup of SCC. In this empirical method, the on-site shear strength of SCC is measured by either the inclined plane (IP) test or the portable vane (PV) test. The static yield stress measured from IPτ 0res@15 min and PVτ 0res@15 min is measured after 15 min of rest, indicating the structural buildup of the concrete at rest. These values are used to calculate Pmax with the pressure envelope being hydrostatic from

Rheology and self-compacting concrete  203

the free surface to Pmax. Khayat et al. (2011) carried out a comprehensive testing program to evaluate the key mixture parameters affecting the formwork pressure exerted by SCC. The investigated parameters included mixture proportions, concrete constituents, concrete temperature, casting characteristics, and the minimum formwork dimension. The UofS2 pressure column and the empirical test methods were employed to evaluate the lateral ­pressure characteristics and related them to the SCC rheological properties. Approximately 780 data points were used to derive equations for predicting from pressures exerted by SCC, employing analyses of multiple parameters using special statistical software. The equations used to calculate the Pmax are as follows:

  

  

Pmax =

Pmax =

wh  [112.5 − 3.8h + 0.6R − 0.6T + 10Dmin − 0.021PVτ 0rest @15min@ T =22°C ] × fMSA  × f WP 100 (8.5)

wh   112 − 3.83h + 0.6R − 0.6T + 10Dmin − 0.023IPτ 0rest @15 min@ T =22°C  × fMSA  × f WP 100  (8.6)

where w is the unit weight of concrete (kN/m3), h is the height of placement (m), R is the rate of placement (m/h), T is the actual concrete temperature (°C), and Dmin is the equivalent to the minimum formwork dimension (d). Use Dmin = d, in case of 0.2